Datasets:

indahl
/

model-data-pack

+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+   1. Definitions.
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+   END OF TERMS AND CONDITIONS
+   APPENDIX: How to apply the Apache License to your work.
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+   Copyright [yyyy] [name of copyright owner]
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+       http://www.apache.org/licenses/LICENSE-2.0
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

CITATION.cff ADDED Viewed

	@@ -0,0 +1,11 @@

+cff-version: 1.2.0
+message: "If you use this data pack, please cite it. It includes MLPerf results published by MLCommons Association (Apache-2.0) and sentences quoted from the sources named in each row. Terms are given by file in LICENSE.txt."
+comment: "Terms are given by file in LICENSE.txt."
+title: "MODEL Data Pack: vendor statements held against published MLCommons results"
+authors:
+  - name: "Heuremen LLC"
+version: "1.0.0"
+date-released: "2026-11-01"
+url: "https://indahl.ai/hermes/model"
+repository-code: "https://github.com/Heuremen-LLC/indahl.ai"
+keywords: [MLPerf, MLCommons, AI hardware, vendor statements, inference, training, storage]

LICENSE.txt ADDED Viewed

	@@ -0,0 +1,47 @@

+MODEL Data Pack
+Heuremen LLC · https://indahl.ai/hermes/model
+This pack mixes data from several sources. Each part keeps the terms of its source, and Heuremen LLC adds no restriction to any of them. The parts are listed here by file.
+1. MLCommons results (Apache License 2.0)
+MLPerf® results published by MLCommons Association, retrieved 28 October 2026 from the repositories named in NOTICE.txt.
+Licence: Apache License 2.0. https://www.apache.org/licenses/LICENSE-2.0 . The text is in APACHE-2.0.txt.
+This applies to every file made from those results:
+- json/accelerators_summary.json
+- json/storage_v30_results.json
+- csv/accelerators.csv
+- csv/accelerators_summary.csv
+- csv/storage_v30_rows.csv
+- csv/storage_v30_meta.csv
+- csv/claims_mlcommons_files.csv
+Changes made by Heuremen LLC: accelerators.csv keeps only closed-division datacenter results that have an accelerator count and a throughput unit, and adds the column per_accelerator, which is the result divided by the number of accelerators in the entry. accelerators_summary.json and its CSV give the count, minimum, median, maximum and best entry for each group. storage_v30_results.json is our aggregation of the Storage v3.0 result files into one row per system and workload, where for training the value is the mean of the runs after the first.
+MLPerf and MLCOMMONS are trademarks of MLCommons Association. The per-accelerator values, ratios and percentages are our own computations and are not verified by MLCommons Association. The trademark and verification lines are in NOTICE.txt. Heuremen LLC is a member of MLCommons, and nothing in this pack is endorsed by MLCommons Association.
+2. Quoted vendor and other sentences
+The sentences in the statement column of the claims files are quoted from the pages and documents named in the source_ columns of the same row. They belong to the owners of those pages, who keep their terms. Each row names its source. To correct or remove a row, use https://indahl.ai/correction.
+This applies to these files, whose columns compared, in_file, outcome and note are Heuremen LLC's own text as item 3 says:
+- json/claims.json
+- csv/claims.csv
+- csv/claims_sources.csv
+3. Heuremen LLC's own parts
+The README, SOURCES.md, NOTICE.txt, LICENSE.txt, TESTED.md, START_HERE.html, verify.py, the build code, the compilation (the flattening, the CSV conversion, the column names), the columns compared, in_file, outcome and note of the claims files, and these files:
+- json/mac_run.jsonl
+- csv/mac_run.csv
+- csv/claims_tally.csv
+- csv/claims_vendor_page_hash.csv
+- csv/claims_meta.csv
+Heuremen LLC permits anyone to use, copy and change these parts with attribution to Heuremen LLC and https://indahl.ai/hermes/model. For its own text (the README, SOURCES.md, NOTICE.txt, LICENSE.txt, TESTED.md, START_HERE.html and the columns compared, in_file, outcome and note of the claims files) that permission does not include use as AI training data.
+Heuremen LLC reserves text and data mining and use as AI training data for that text only. That reservation does not apply to the MLCommons results or to the sentences quoted from the sources.

NOTICE.txt ADDED Viewed

	@@ -0,0 +1,40 @@

+MODEL Data Pack
+Heuremen LLC · https://indahl.ai/hermes/model
+Attribution. The full terms are in LICENSE.txt and this notice says the same.
+MLCommons results
+  MLPerf® Inference v5.0, v5.1, v6.0 and v6.1 results, datacenter and edge suites, closed and open divisions, as published by MLCommons Association in summary_results.json of the repositories inference_results_v5.0, inference_results_v5.1 and inference_results_v6.0 and in summary.json of the repository inference_results_v6.1 at github.com/mlcommons, retrieved 28 October 2026. MLPerf® Training v5.0, v5.1 and v6.0 results as published in the result logs of the repositories training_results_v5.0, training_results_v5.1 and training_results_v6.0, retrieved 28 October 2026. MLPerf® Storage v3.0 results as published in the repository storage_results_v3.0, retrieved 28 October 2026.
+  Licence: Apache License 2.0. https://www.apache.org/licenses/LICENSE-2.0 . The text is in APACHE-2.0.txt.
+  Applies to these files:
+  - json/accelerators_summary.json
+  - json/storage_v30_results.json
+  - csv/accelerators.csv
+  - csv/accelerators_summary.csv
+  - csv/storage_v30_rows.csv
+  - csv/storage_v30_meta.csv
+  - csv/claims_mlcommons_files.csv
+  Changes by Heuremen LLC: accelerators.csv keeps only closed-division datacenter results that have an accelerator count and a throughput unit, and adds the column per_accelerator, which is the result divided by the number of accelerators in the entry. accelerators_summary.json and the CSV made from it give the count, minimum, median, maximum and best entry for each group. storage_v30_results.json is our aggregation of the Storage v3.0 result files into one row per system and workload, where for training the value is the mean of the runs after the first. The per-accelerator values, ratios and percentages are our own computations. They are not primary MLPerf® metrics and are not verified by MLCommons Association.
+Trademarks and verification
+  Unverified results have not been through an MLPerf® review and may use measurement methodologies and/or workload implementations that are inconsistent with the MLPerf® specification for verified results. The per-accelerator values, ratios and percentages in this pack are our own computations from the MLCommons results. Result not verified by MLCommons Association.
+  The MLPerf® and MLCOMMONS® trademarks and design versions thereof are registered and unregistered trademarks and service marks of MLCommons Association in the United States and other countries. All rights reserved. Unauthorized use is strictly prohibited. See www.mlcommons.org for more information.
+  Heuremen LLC is a member of MLCommons, and nothing in this pack is endorsed by MLCommons Association.
+  Where this pack names an MLPerf result, the row carries the benchmark suite and version, the model and scenario, and the MLPerf result ID, and the date and source of retrieval are given above. The accelerators.csv rows are Closed division, datacenter suite. MLPerf results may only be compared with compatible MLPerf results, and the MLPerf Results Messaging Guidelines at https://github.com/mlcommons/policies/blob/master/MLPerf_Results_Messaging_Guidelines.adoc apply to anyone who repeats them.
+Vendor and other sources
+  Sentences in the statement column of csv/claims.csv and in json/claims.json are quoted from the page or document named in the source_ columns of the same row, fetched on 2026-10-28. They belong to the owners of those pages and keep their terms. They are quoted so that each statement can be set beside the published results.
+  Applies to these files:
+  - json/claims.json
+  - csv/claims.csv
+  - csv/claims_sources.csv
+  The columns compared, in_file, outcome and note of these files are Heuremen LLC's own text.
+Heuremen LLC
+  The README, SOURCES.md, NOTICE.txt, LICENSE.txt, TESTED.md, START_HERE.html, verify.py, the build code, the compilation and the CSV conversion, the readings in json/mac_run.jsonl and csv/mac_run.csv, and these files may be used, copied and changed with attribution to Heuremen LLC and https://indahl.ai/hermes/model:
+  - json/mac_run.jsonl
+  - csv/mac_run.csv
+  - csv/claims_tally.csv
+  - csv/claims_vendor_page_hash.csv
+  - csv/claims_meta.csv
+  For the text files (the README, SOURCES.md, NOTICE.txt, LICENSE.txt, TESTED.md and START_HERE.html) and for the columns compared, in_file, outcome and note of the claims files, that permission does not include use as AI training data. The reservation against text and data mining and AI training covers that text only. It does not apply to the MLCommons results or to the sentences quoted from the sources.

README.md ADDED Viewed

	@@ -0,0 +1,100 @@

+---
+pretty_name: "MODEL Data Pack"
+license: other
+license_name: apache-2.0-for-mlcommons-results-see-license-txt
+license_link: https://huggingface.co/datasets/indahl/model-data-pack/blob/main/LICENSE.txt
+tags:
+- mlperf
+- mlcommons
+- hardware
+- benchmarks
+- reproducibility
+configs:
+- config_name: claims
+  default: true
+  data_files: csv/claims.csv
+- config_name: claims_sources
+  data_files: csv/claims_sources.csv
+- config_name: claims_tally
+  data_files: csv/claims_tally.csv
+- config_name: accelerators
+  data_files: csv/accelerators.csv
+- config_name: accelerators_summary
+  data_files: csv/accelerators_summary.csv
+- config_name: storage_v30_rows
+  data_files: csv/storage_v30_rows.csv
+---
+# MODEL Data Pack
+This is the same data as the download on indahl.ai, published here for the Hugging Face viewer and the datasets library. The page is https://indahl.ai/hermes/model . The files were tested before release, and what was opened and compared is in [TESTED.md](https://huggingface.co/datasets/indahl/model-data-pack/blob/main/TESTED.md). The checking script `verify.py` and the guide page `START_HERE.html` are in the download from indahl.ai and are not copied here. Licences differ by file. MLCommons-derived files are under the Apache License 2.0 and the rest follow the pack LICENSE.txt. Full terms: [LICENSE.txt](https://huggingface.co/datasets/indahl/model-data-pack/blob/main/LICENSE.txt).
+Heuremen LLC · https://indahl.ai/hermes/model
+The data behind the MODEL page at https://indahl.ai/hermes/model, as files you can open in a spreadsheet or Python. MODEL sets what AI hardware vendors state about their systems next to the results MLCommons publishes. Each file comes as the page publishes it, and each JSON file also comes as CSV with one row per record. The statements were fetched and compared on 2026-10-28, the files were built on 2026-10-28, and the MLCommons files were retrieved on 28 October 2026. Vendor pages and MLCommons files change, so a later read can differ.
+## What is in each file
+Statements and outcomes (`json/claims.json`): 531 statements from 74 sources, each with an outcome of ABOVE, BELOW, CONSISTENT or CANNOT_TELL. The file has a `summary` entry and a `claims` list. The rows come as `csv/claims.csv`. The columns are `id`, `vendor`, `source_key`, `statement`, `verbatim`, `compared`, `stated`, `in_file`, `outcome`, `note`, `label_mismatch` and four columns that start with `source_`, which hold the title, address, fetch date and round of the page the sentence comes from. The `id` repeats between sources, so a row is named by `source_key` and `id` together. The `statement` column holds the sentence as published, without quotation marks. The page puts it in quotation marks when `verbatim` is true. The `compared` column names the MLCommons entries the statement points to, such as 5.1-0072, and the `in_file` column gives what the files show. In 2 rows `label_mismatch` is true, and the `in_file` text of those rows says how the label in the file differs from the label in the statement. For those rows the page adds one sentence to the note, which reads: The entry named in the post is listed in the file with a different accelerator or model.
+The summary comes as five more files. `csv/claims_sources.csv` has one row per source with its vendor, title, address and round. `csv/claims_tally.csv` has the number of statements per source and outcome, and a blank means none. `csv/claims_vendor_page_hash.csv` has the SHA-256 of the vendor page as we read it. `csv/claims_mlcommons_files.csv` has the address, SHA-256 and row count of each MLCommons inference file we used. `csv/claims_meta.csv` has the remaining summary fields as key and value, such as the build date and the commits of the training and storage repositories.
+Results by accelerator (`csv/accelerators.csv` and `json/accelerators_summary.json`): 1,727 closed-division datacenter throughput results from MLPerf Inference v5.0, v5.1 and v6.0 that have an accelerator count, in 569 groups of version, model, scenario and accelerator. The `per_accelerator` column is `result` divided by `total`, which is our own computation and not an MLPerf primary metric. The summary file gives the number of submissions and submitters, the minimum, median and maximum, and the entry with the highest value for each group. `csv/accelerators_summary.csv` holds the same groups as CSV. The page states which rows are left out and why.
+Storage results (`json/storage_v30_results.json`): 144 rows that aggregate the MLPerf Storage v3.0 result files into one row per system and workload, where for training the value is the mean of the runs after the first, as `csv/storage_v30_rows.csv`, with the source and licence in `csv/storage_v30_meta.csv`. Rows for different workloads have different measurement columns, so a column is blank where the workload has no such measurement.
+Readings from our unit (`json/mac_run.jsonl`): 17 lines of memory bandwidth readings from the MacBook Pro 14-inch with M5 Max that the Apple table on the page uses. Each line has a `kind`. They come as `csv/mac_run.csv`, where a column is blank for a line of another kind.
+## Sources and dates
+SOURCES.md lists the upstream sources, their links, their terms and the dates found in the data. NOTICE.txt carries the MLCommons attribution and trademark lines. The statements were fetched on 2026-10-28. The MLCommons files were retrieved on 28 October 2026. Every row of `csv/claims.csv` carries the address of the page its sentence comes from and the date it was fetched.
+## How to open the files
+Spreadsheet: open any file in `csv/`. The files are UTF-8, and the ones we wrote carry a byte order mark so Excel reads names correctly. Empty text and missing values are both shown as blank cells. The file `csv/accelerators.csv` is the published file without a byte order mark.
+Python, standard library only:
+```
+import json, csv
+claims = json.load(open("json/claims.json", encoding="utf-8"))["claims"]
+rows = list(csv.DictReader(open("csv/claims.csv", encoding="utf-8-sig")))
+acc = list(csv.DictReader(open("csv/accelerators.csv", encoding="utf-8-sig")))
+```
+Check the pack with `python verify.py`. It recounts the rows of every file, compares every CSV cell with the JSON, recomputes the outcome counts of each source from the statements, and recomputes every group in `accelerators_summary.json` from the rows of `accelerators.csv`.
+## What is here
+- `json/`: the original files from the MODEL data folder, unchanged
+- `csv/`: one CSV for each table, plus the `_meta.csv` files, and the published `accelerators.csv`
+- `SOURCES.md`: upstream sources, links, terms and dates
+- `TESTED.md`: what the files were opened with and what was compared
+- `NOTICE.txt`: MLCommons attribution and trademark lines, and the terms of each source
+- `LICENSE.txt`: terms by file
+- `APACHE-2.0.txt`: the Apache License 2.0, which covers the MLCommons results
+- `verify.py`: recounts and compares the files
+- `START_HERE.html`: this guide as a page
+- `CITATION.cff`
+## Links
+- MODEL: https://indahl.ai/hermes/model
+- Papers table: https://indahl.ai/hermes/model/papers
+- Code: https://github.com/Heuremen-LLC/indahl.ai
+- MLCommons: https://mlcommons.org
+- MLPerf results messaging guidelines: https://github.com/mlcommons/policies/blob/master/MLPerf_Results_Messaging_Guidelines.adoc
+- Correction or removal form: https://indahl.ai/correction
+## Notes on the data
+The files in `json/` are the files the MODEL page links, unchanged. The page also links the scripts that recompute each row from the MLCommons files, and those scripts are not in this pack. They are in the code folder at https://github.com/Heuremen-LLC/indahl.ai.
+The papers table on the MODEL papers page and the live model-claim rounds are not in this pack. The papers table has no data file, and the live rounds are a feed that the page reloads every 30 seconds.
+## Terms
+Each part of this pack keeps the terms of its source, and LICENSE.txt gives them by file. The MLCommons results are published by MLCommons Association under the Apache License 2.0, and the Apache text is included. MLPerf and MLCOMMONS are trademarks of MLCommons Association. The per-accelerator values, ratios and percentages here are our own computations and are not verified by MLCommons Association. The sentences quoted from vendor pages belong to the pages named in each row. To correct or remove a row, use https://indahl.ai/correction. It is open to anyone, and we acknowledge every request within three business days.
+Heuremen LLC permits anyone to use, copy and change its own parts with attribution to Heuremen LLC and https://indahl.ai/hermes/model. For its own text (the README, SOURCES.md, NOTICE.txt, LICENSE.txt, TESTED.md, START_HERE.html and the `compared`, `in_file`, `outcome` and `note` columns of the claims files) that permission does not include use as AI training data. Its reservation against text and data mining and AI training covers that text only and does not apply to the MLCommons results or to the sentences quoted from the sources.

SOURCES.md ADDED Viewed

	@@ -0,0 +1,30 @@

+# Sources
+Heuremen LLC · https://indahl.ai/hermes/model
+Each entry gives the upstream source as the MODEL page and the data files state it, with the dates found in the data. The descriptions follow https://indahl.ai/hermes/model.
+## MLCommons results
+- MLPerf Inference v5.0, v5.1 and v6.0, datacenter and edge suites, closed and open divisions, as published by MLCommons Association in `summary_results.json` of the repositories `inference_results_v5.0`, `inference_results_v5.1` and `inference_results_v6.0` at https://github.com/mlcommons, retrieved 28 October 2026. Terms: Apache License 2.0, as the page states. The repository `inference_results_v5.0` has no LICENSE file in its root on 2026-11-01, and the page states Apache-2.0 for it. The other repositories listed here have a LICENSE.md with the Apache License 2.0.
+- MLPerf Inference v6.1 in `summary.json` of the repository `inference_results_v6.1`, retrieved 28 October 2026. Terms: Apache License 2.0.
+- Rows and SHA-256 of the files we used, as the page states them and as `csv/claims_mlcommons_files.csv` gives them in full: v5.0 17,454 rows, v5.1 1,448 rows, v6.0 520 rows, v6.1 492 rows.
+- MLPerf Training v5.0, v5.1 and v6.0, as published in the result logs of the repositories `training_results_v5.0`, `training_results_v5.1` and `training_results_v6.0` at https://github.com/mlcommons, retrieved 28 October 2026. Terms: Apache License 2.0. The commits and the numbers of result sets are in `csv/claims_meta.csv`. Training times on the page are recomputed from run start and run stop in each log, with the olympic average and review-committee scaling factors that the MLCommons result summarizer applies.
+- MLPerf Storage v3.0, as published in the repository `storage_results_v3.0` at https://github.com/mlcommons/storage_results_v3.0, retrieved 28 October 2026. Terms: Apache License 2.0, as `json/storage_v30_results.json` states in its `licence` entry. The commit and the number of result rows are in `csv/claims_meta.csv`.
+- How results may be reused: the MLPerf Results Messaging Guidelines at https://github.com/mlcommons/policies/blob/master/MLPerf_Results_Messaging_Guidelines.adoc and the policies listed at https://mlcommons.org/policies/. NOTICE.txt gives the attribution and trademark lines.
+## Vendor and other statements (json/claims.json)
+- Each row of `csv/claims.csv` quotes a sentence from one source page or document. The title, address, fetch date and round of that source are in the `source_` columns of the row and in `csv/claims_sources.csv`. All sources were fetched on 2026-10-28. The SHA-256 of each vendor page as we read it is in `csv/claims_vendor_page_hash.csv`.
+- The sources are web pages and PDFs of the vendors and of MLCommons, and one source is the test output of a software project. The Apple table compares the Apple specification page with readings from our own unit.
+- The sentences belong to the owners of those pages. They are quoted so that each statement can be set beside the published results, and each row names its source. The terms of each page are its owner's.
+- The columns `compared`, `in_file`, `outcome` and `note` are ours. The `compared` column names the MLCommons entries, and the `in_file` numbers are computed by us from the public MLCommons files.
+## Results by accelerator (csv/accelerators.csv)
+- Computed by us from `summary_results.json` of MLPerf Inference v5.0, v5.1 and v6.0 (Apache License 2.0). Closed-division datacenter results with an accelerator count and a throughput unit (tokens, samples or queries per second). The per-accelerator value is the result divided by the number of accelerators in the entry. Where an entry's accelerator name states its own count and the count differs from nodes times accelerators per node, the count in the name is used. The page lists the rows left out: no accelerator count 49, 61 and 24, and given as a latency 6, 0 and 7, for v5.0, v5.1 and v6.0 in that order.
+- Per-accelerator throughput is not an MLPerf primary metric.
+## Readings from our unit (json/mac_run.jsonl)
+- Memory bandwidth readings taken by us on 2026-10-28 on a MacBook Pro 14-inch with M5 Max. These are our own data. The scripts that took them are linked from the MODEL page.

TESTED.md ADDED Viewed

	@@ -0,0 +1,28 @@

+# Tested
+Heuremen LLC · https://indahl.ai/hermes/model
+On 2026-11-01 the zip was unzipped into an empty folder and every CSV was opened with pandas 3.1.0 on Python 3.14 at default settings. The live MODEL page and its 74 source pages were fetched the same day and read with lxml 6.1 and Beautiful Soup 4. Every count and number below was compared with the live pages and matched, except the one note named further down.
+- Live files: claims.json, accelerators_summary.json, storage_v30_results.json, mac_run.jsonl and accelerators.csv are byte for byte the files in the pack
+- Statements: 531 rows in 74 sources, with 8 ABOVE, 51 BELOW, 368 CONSISTENT and 104 CANNOT_TELL, which are the sums of the source table on the page
+- Source table: vendor, round and the four outcome counts of all 74 rows equal `claims.csv` recounted with pandas and `claims_tally.csv`
+- Source pages: the sentence, compared, stated, in the file, outcome and note of all 531 statements, and the line N statements checked of each of the 74 pages
+- Results by accelerator: 1,727 rows in 569 groups, 814 in v5.0, 493 in v5.1 and 420 in v6.0, and `per_accelerator` equals `result` divided by `total` in every row
+- Left out rows: no accelerator count 49, 61 and 24, and latency units 6, 0 and 7, for v5.0, v5.1 and v6.0
+- MLCommons files: rows 17,454, 1,448, 520 and 492 for v5.0, v5.1, v6.0 and v6.1, with the first 16 characters of each SHA-256 as the page prints them
+- Training and storage: result sets 229, 201 and 185 with commits eabf23a07b2a, d7e703fccc1c and 441d65f4445c, and 144 storage result rows with commit 802035fe6580
+- Known answer: entry 5.0-0001, llama2-70b-99, Offline, 33,928.1 tokens per second on 8 accelerators is 4,241.0125 per accelerator
+- Known answer: the statement NVIDIA v5.1 C1 reads +45.2% and CONSISTENT, and 5.1-0072 over 5.1-0071 for DeepSeek-R1 Offline recomputes from `accelerators.csv` as 1.452
+The id column repeats between sources, with 500 different values over 531 rows, so a row is named by `source_key` and `id` together, and that pair is unique.
+The test found one note, Nebius Storage v3.0 with `id` NB11, worded differently in the published `claims.json` than on the page. The published file was corrected on 2 November 2026 to match the page, and this pack carries the corrected file.
+For the rows with `label_mismatch` true, the page adds the sentence The entry named in the post is listed in the file with a different accelerator or model. to the note. The pack does not repeat it in the note column.
+At default settings pandas reads a blank cell as missing. In `csv/claims_tally.csv` a blank means no statement with that outcome. No cell in any file is text that pandas reads as missing, such as NA or None.
+The file `verify.py` compares every CSV cell with the JSON, recomputes the outcome counts of each source from the statements, and recomputes every group of `accelerators_summary.json` from `accelerators.csv`.
+Not tested: Excel or LibreOffice, the scripts that recompute the rows from the MLCommons files, and a new download of the MLCommons files. The page prints the first 16 characters of each SHA-256 and the pack holds the full values, and those were compared, but the files were not downloaded again.

csv/accelerators.csv ADDED Viewed

The diff for this file is too large to render. See raw diff

csv/accelerators_summary.csv ADDED Viewed

	@@ -0,0 +1,570 @@

+version,model,scenario,accelerator,units,submissions,submitters,min,median,max,best_id
+v5.0,3d-unet-99,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,4.14253,4.14253,4.14253,5.0-0024
+v5.0,3d-unet-99,Offline,NVIDIA H100-NVL-94GB,Samples/s,7,5,5.3934,5.41895,5.60685,5.0-0050
+v5.0,3d-unet-99,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,4.63505,4.63505,4.63505,5.0-0064
+v5.0,3d-unet-99,Offline,NVIDIA H100-SXM-80GB,Samples/s,4,3,6.418775,6.47814375,6.522075,5.0-0039
+v5.0,3d-unet-99,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,6.174525,6.184,6.1929875,5.0-0004
+v5.0,3d-unet-99,Offline,NVIDIA H200-SXM-141GB,Samples/s,11,7,6.821175,6.8341125,6.8874375,5.0-0041
+v5.0,3d-unet-99,Offline,NVIDIA L40S,Samples/s,4,3,3.3889,3.67361125,3.8778,5.0-0012
+v5.0,3d-unet-99,Offline,Virtualized NVIDIA H100-SXM-80GB,Samples/s,1,1,6.4223,6.4223,6.4223,5.0-0007
+v5.0,3d-unet-99.9,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,4.15438,4.15438,4.15438,5.0-0024
+v5.0,3d-unet-99.9,Offline,NVIDIA H100-NVL-94GB,Samples/s,5,4,5.4049,5.4772,5.60685,5.0-0050
+v5.0,3d-unet-99.9,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,4.63505,4.63505,4.63505,5.0-0064
+v5.0,3d-unet-99.9,Offline,NVIDIA H100-SXM-80GB,Samples/s,4,3,6.418775,6.47925625,6.525425,5.0-0040
+v5.0,3d-unet-99.9,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,6.174525,6.184,6.1929875,5.0-0004
+v5.0,3d-unet-99.9,Offline,NVIDIA H200-SXM-141GB,Samples/s,11,7,6.821175,6.8341125,6.89035,5.0-0041
+v5.0,3d-unet-99.9,Offline,NVIDIA L40S,Samples/s,2,2,3.3889,3.4460687500000002,3.5032375,5.0-0018
+v5.0,3d-unet-99.9,Offline,Virtualized NVIDIA H100-SXM-80GB,Samples/s,1,1,6.4223,6.4223,6.4223,5.0-0007
+v5.0,bert-99,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,4133.095,4133.095,4133.095,5.0-0024
+v5.0,bert-99.9,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,1669.855,1669.855,1669.855,5.0-0024
+v5.0,dlrm-v2-99,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Samples/s,3,2,85998.9,86971.5,87268.0,5.0-0059
+v5.0,dlrm-v2-99,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Samples/s,2,2,83504.2,84745.54999999999,85986.9,5.0-0069
+v5.0,dlrm-v2-99,Offline,NVIDIA H100-NVL-94GB,Samples/s,2,1,50884.0,51203.75,51523.5,5.0-0009
+v5.0,dlrm-v2-99,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,46766.0,46766.0,46766.0,5.0-0064
+v5.0,dlrm-v2-99,Offline,NVIDIA H100-SXM-80GB,Samples/s,2,2,73274.25,74086.5625,74898.875,5.0-0039
+v5.0,dlrm-v2-99,Offline,NVIDIA H200-NVL-141GB,Samples/s,2,2,40833.75,52439.5625,64045.375,5.0-0043
+v5.0,dlrm-v2-99,Offline,NVIDIA H200-SXM-141GB,Samples/s,6,6,78883.0,80336.5625,81811.125,5.0-0041
+v5.0,dlrm-v2-99,Offline,NVIDIA L40S,Samples/s,3,2,28162.125,28916.25,28959.05,5.0-0013
+v5.0,dlrm-v2-99,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Queries/s,3,2,78980.5,80994.2,80994.3,5.0-0037
+v5.0,dlrm-v2-99,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Queries/s,2,2,80992.9,81892.85,82792.8,5.0-0066
+v5.0,dlrm-v2-99,Server,NVIDIA H100-NVL-94GB,Queries/s,2,1,48493.9,48893.4,49292.9,5.0-0014
+v5.0,dlrm-v2-99,Server,NVIDIA H100-PCIe-80GB,Queries/s,1,1,43854.75,43854.75,43854.75,5.0-0064
+v5.0,dlrm-v2-99,Server,NVIDIA H100-SXM-80GB,Queries/s,2,2,64015.125,66734.5,69453.875,5.0-0039
+v5.0,dlrm-v2-99,Server,NVIDIA H200-NVL-141GB,Queries/s,2,2,25126.375,45383.0625,65639.75,5.0-0003
+v5.0,dlrm-v2-99,Server,NVIDIA H200-SXM-141GB,Queries/s,6,6,69485.375,73145.5,73770.875,5.0-0041
+v5.0,dlrm-v2-99,Server,NVIDIA L40S,Queries/s,3,2,22205.375,24795.8,24795.8,5.0-0012
+v5.0,dlrm-v2-99.9,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Samples/s,3,2,51970.0,51974.5,52549.1,5.0-0059
+v5.0,dlrm-v2-99.9,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Samples/s,2,2,50606.5,50818.7,51030.9,5.0-0069
+v5.0,dlrm-v2-99.9,Offline,NVIDIA H100-NVL-94GB,Samples/s,1,1,28205.0,28205.0,28205.0,5.0-0044
+v5.0,dlrm-v2-99.9,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,27352.0,27352.0,27352.0,5.0-0064
+v5.0,dlrm-v2-99.9,Offline,NVIDIA H100-SXM-80GB,Samples/s,2,2,46579.0,46871.0625,47163.125,5.0-0039
+v5.0,dlrm-v2-99.9,Offline,NVIDIA H200-NVL-141GB,Samples/s,2,2,40035.875,40381.875,40727.875,5.0-0003
+v5.0,dlrm-v2-99.9,Offline,NVIDIA H200-SXM-141GB,Samples/s,5,5,47879.875,48321.125,49770.5,5.0-0041
+v5.0,dlrm-v2-99.9,Offline,NVIDIA L40S,Samples/s,1,1,12564.625,12564.625,12564.625,5.0-0045
+v5.0,dlrm-v2-99.9,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Queries/s,3,2,49081.05,50070.5,50991.7,5.0-0059
+v5.0,dlrm-v2-99.9,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Queries/s,2,2,45990.4,48229.65,50468.9,5.0-0069
+v5.0,dlrm-v2-99.9,Server,NVIDIA H100-NVL-94GB,Queries/s,1,1,21252.25,21252.25,21252.25,5.0-0044
+v5.0,dlrm-v2-99.9,Server,NVIDIA H100-PCIe-80GB,Queries/s,1,1,25046.75,25046.75,25046.75,5.0-0064
+v5.0,dlrm-v2-99.9,Server,NVIDIA H100-SXM-80GB,Queries/s,2,2,42505.5,43631.1875,44756.875,5.0-0039
+v5.0,dlrm-v2-99.9,Server,NVIDIA H200-NVL-141GB,Queries/s,2,2,24439.75,30971.6875,37503.625,5.0-0003
+v5.0,dlrm-v2-99.9,Server,NVIDIA H200-SXM-141GB,Queries/s,5,5,45331.875,46256.875,47507.375,5.0-0041
+v5.0,dlrm-v2-99.9,Server,NVIDIA L40S,Queries/s,1,1,11873.6625,11873.6625,11873.6625,5.0-0045
+v5.0,gptj-99,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,2824.83,2833.08,2845.43,5.0-0037
+v5.0,gptj-99,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,2808.21,2808.21,2808.21,5.0-0066
+v5.0,gptj-99,Offline,NVIDIA H100-NVL-94GB,Tokens/s,5,4,1810.035,1871.68,1926.4925,5.0-0050
+v5.0,gptj-99,Offline,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1579.7675,1579.7675,1579.7675,5.0-0064
+v5.0,gptj-99,Offline,NVIDIA H100-SXM-80GB,Tokens/s,5,4,2538.4625,2582.9625,2624.2375,5.0-0039
+v5.0,gptj-99,Offline,NVIDIA H200-NVL-141GB,Tokens/s,4,4,2112.5125,2249.76875,2340.525,5.0-0071
+v5.0,gptj-99,Offline,NVIDIA H200-SXM-141GB,Tokens/s,16,12,2624.95,2670.73125,2706.375,5.0-0052
+v5.0,gptj-99,Offline,NVIDIA L40S,Tokens/s,3,3,852.98875,868.335,877.485,5.0-0013
+v5.0,gptj-99,Offline,Virtualized NVIDIA H100-SXM-80GB,Tokens/s,1,1,2492.0625,2492.0625,2492.0625,5.0-0007
+v5.0,gptj-99,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,2519.91,2526.03,2526.75,5.0-0059
+v5.0,gptj-99,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,2824.55,2824.55,2824.55,5.0-0066
+v5.0,gptj-99,Server,NVIDIA H100-NVL-94GB,Tokens/s,5,4,1652.635,1794.965,1855.8375,5.0-0023
+v5.0,gptj-99,Server,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1562.8175,1562.8175,1562.8175,5.0-0064
+v5.0,gptj-99,Server,NVIDIA H100-SXM-80GB,Tokens/s,5,4,2414.6125,2588.3375,2641.8,5.0-0020
+v5.0,gptj-99,Server,NVIDIA H200-NVL-141GB,Tokens/s,4,4,2244.2375,2252.5562499999996,2321.3875,5.0-0071
+v5.0,gptj-99,Server,NVIDIA H200-SXM-141GB,Tokens/s,16,12,2562.725,2692.51875,2726.675,5.0-0011
+v5.0,gptj-99,Server,NVIDIA L40S,Tokens/s,3,3,776.6675,852.63875,872.89,5.0-0013
+v5.0,gptj-99,Server,Virtualized NVIDIA H100-SXM-80GB,Tokens/s,1,1,2414.8625,2414.8625,2414.8625,5.0-0007
+v5.0,gptj-99.9,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,2824.83,2833.08,2845.43,5.0-0037
+v5.0,gptj-99.9,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,2808.21,2808.21,2808.21,5.0-0066
+v5.0,gptj-99.9,Offline,NVIDIA H100-NVL-94GB,Tokens/s,2,2,1846.2125,1886.3525,1926.4925,5.0-0050
+v5.0,gptj-99.9,Offline,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1579.7675,1579.7675,1579.7675,5.0-0064
+v5.0,gptj-99.9,Offline,NVIDIA H100-SXM-80GB,Tokens/s,5,4,2538.4625,2582.9625,2599.5,5.0-0039
+v5.0,gptj-99.9,Offline,NVIDIA H200-NVL-141GB,Tokens/s,4,4,2112.5125,2249.76875,2340.525,5.0-0071
+v5.0,gptj-99.9,Offline,NVIDIA H200-SXM-141GB,Tokens/s,16,12,2624.95,2669.44375,2706.375,5.0-0052
+v5.0,gptj-99.9,Offline,NVIDIA L40S,Tokens/s,2,2,852.98875,860.661875,868.335,5.0-0018
+v5.0,gptj-99.9,Offline,Virtualized NVIDIA H100-SXM-80GB,Tokens/s,1,1,2492.0625,2492.0625,2492.0625,5.0-0007
+v5.0,gptj-99.9,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,2519.91,2526.03,2526.75,5.0-0059
+v5.0,gptj-99.9,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,2824.55,2824.55,2824.55,5.0-0066
+v5.0,gptj-99.9,Server,NVIDIA H100-NVL-94GB,Tokens/s,2,2,1693.4675,1774.6525000000001,1855.8375,5.0-0023
+v5.0,gptj-99.9,Server,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1562.8175,1562.8175,1562.8175,5.0-0064
+v5.0,gptj-99.9,Server,NVIDIA H100-SXM-80GB,Tokens/s,5,4,2414.6125,2588.3375,2641.8,5.0-0020
+v5.0,gptj-99.9,Server,NVIDIA H200-NVL-141GB,Tokens/s,4,4,2244.2375,2252.5562499999996,2321.3875,5.0-0071
+v5.0,gptj-99.9,Server,NVIDIA H200-SXM-141GB,Tokens/s,16,12,2562.725,2692.51875,2726.675,5.0-0011
+v5.0,gptj-99.9,Server,NVIDIA L40S,Tokens/s,2,2,776.6675,814.653125,852.63875,5.0-0045
+v5.0,gptj-99.9,Server,Virtualized NVIDIA H100-SXM-80GB,Tokens/s,1,1,2414.8625,2414.8625,2414.8625,5.0-0007
+v5.0,llama2-70b-99,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,1,1,3224.4375,3224.4375,3224.4375,5.0-0054
+v5.0,llama2-70b-99,Offline,AMD Instinct MI325X 256GB HBM3E,Tokens/s,4,4,4133.6375,4163.6125,4241.0125,5.0-0001
+v5.0,llama2-70b-99,Offline,NVIDIA B200-SXM-180GB,Tokens/s,5,3,12140.5,12269.3,12357.25,5.0-0056
+v5.0,llama2-70b-99,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,4609.05,4681.425,4686.14,5.0-0059
+v5.0,llama2-70b-99,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,3144.18,3144.18,3144.18,5.0-0066
+v5.0,llama2-70b-99,Offline,NVIDIA H100-NVL-94GB,Tokens/s,3,2,1870.66,1939.975,2260.6125,5.0-0023
+v5.0,llama2-70b-99,Offline,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1753.2925,1753.2925,1753.2925,5.0-0064
+v5.0,llama2-70b-99,Offline,NVIDIA H100-SXM-80GB,Tokens/s,6,5,3862.4625,3899.46875,3913.35,5.0-0057
+v5.0,llama2-70b-99,Offline,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3470.6625,3787.1875,3893.7375,5.0-0017
+v5.0,llama2-70b-99,Offline,NVIDIA H200-SXM-141GB,Tokens/s,18,13,4222.725,4348.143749999999,4431.625,5.0-0051
+v5.0,llama2-70b-99,Offline,NVIDIA L40S,Tokens/s,2,2,435.19125,446.08875,456.98625,5.0-0046
+v5.0,llama2-70b-99,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,1,1,2907.496875,2907.496875,2907.496875,5.0-0054
+v5.0,llama2-70b-99,Server,AMD Instinct MI325X 256GB HBM3E,Tokens/s,4,4,3596.275,3698.8,3840.5625,5.0-0001
+v5.0,llama2-70b-99,Server,NVIDIA B200-SXM-180GB,Tokens/s,5,3,11542.3375,12194.0125,12305.4125,5.0-0056
+v5.0,llama2-70b-99,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,4327.68,4337.29,4350.26,5.0-0059
+v5.0,llama2-70b-99,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,3203.87,3203.87,3203.87,5.0-0066
+v5.0,llama2-70b-99,Server,NVIDIA H100-NVL-94GB,Tokens/s,3,2,1668.775,1668.9,2106.075,5.0-0023
+v5.0,llama2-70b-99,Server,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1222.48,1222.48,1222.48,5.0-0064
+v5.0,llama2-70b-99,Server,NVIDIA H100-SXM-80GB,Tokens/s,6,5,3324.925,3809.7875000000004,3888.2875,5.0-0057
+v5.0,llama2-70b-99,Server,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3427.1625,3454.8625,3606.05,5.0-0017
+v5.0,llama2-70b-99,Server,NVIDIA H200-SXM-141GB,Tokens/s,18,13,3663.2,4128.25,4133.9375,5.0-0060
+v5.0,llama2-70b-99,Server,NVIDIA L40S,Tokens/s,2,2,397.70625,398.92375000000004,400.14125,5.0-0018
+v5.0,llama2-70b-99.9,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,1,1,3224.4375,3224.4375,3224.4375,5.0-0054
+v5.0,llama2-70b-99.9,Offline,AMD Instinct MI325X 256GB HBM3E,Tokens/s,4,4,4133.6375,4163.6125,4241.0125,5.0-0001
+v5.0,llama2-70b-99.9,Offline,NVIDIA B200-SXM-180GB,Tokens/s,5,3,12136.2375,12269.3,12357.25,5.0-0056
+v5.0,llama2-70b-99.9,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,4609.05,4681.425,4686.14,5.0-0059
+v5.0,llama2-70b-99.9,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,3144.18,3144.18,3144.18,5.0-0066
+v5.0,llama2-70b-99.9,Offline,NVIDIA H100-NVL-94GB,Tokens/s,1,1,2260.6125,2260.6125,2260.6125,5.0-0023
+v5.0,llama2-70b-99.9,Offline,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1753.2925,1753.2925,1753.2925,5.0-0064
+v5.0,llama2-70b-99.9,Offline,NVIDIA H100-SXM-80GB,Tokens/s,6,5,3862.4625,3897.7625,3913.35,5.0-0057
+v5.0,llama2-70b-99.9,Offline,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3470.6625,3787.1875,3893.7375,5.0-0017
+v5.0,llama2-70b-99.9,Offline,NVIDIA H200-SXM-141GB,Tokens/s,17,12,4245.5,4350.8875,4431.625,5.0-0051
+v5.0,llama2-70b-99.9,Offline,NVIDIA L40S,Tokens/s,2,2,435.19125,446.08875,456.98625,5.0-0046
+v5.0,llama2-70b-99.9,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,1,1,2907.496875,2907.496875,2907.496875,5.0-0054
+v5.0,llama2-70b-99.9,Server,AMD Instinct MI325X 256GB HBM3E,Tokens/s,4,4,3596.275,3698.8,3840.5625,5.0-0001
+v5.0,llama2-70b-99.9,Server,NVIDIA B200-SXM-180GB,Tokens/s,5,3,11525.4625,12194.0125,12305.4125,5.0-0056
+v5.0,llama2-70b-99.9,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,4327.68,4337.29,4350.26,5.0-0059
+v5.0,llama2-70b-99.9,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,1,1,3203.87,3203.87,3203.87,5.0-0066
+v5.0,llama2-70b-99.9,Server,NVIDIA H100-NVL-94GB,Tokens/s,1,1,2106.075,2106.075,2106.075,5.0-0023
+v5.0,llama2-70b-99.9,Server,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,1222.48,1222.48,1222.48,5.0-0064
+v5.0,llama2-70b-99.9,Server,NVIDIA H100-SXM-80GB,Tokens/s,6,5,3324.925,3811.3374999999996,3888.2875,5.0-0057
+v5.0,llama2-70b-99.9,Server,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3427.1625,3454.8625,3606.05,5.0-0017
+v5.0,llama2-70b-99.9,Server,NVIDIA H200-SXM-141GB,Tokens/s,17,12,3663.2,4128.0875,4133.9375,5.0-0060
+v5.0,llama2-70b-99.9,Server,NVIDIA L40S,Tokens/s,2,2,397.70625,398.92375000000004,400.14125,5.0-0018
+v5.0,llama2-70b-interactive-99,Offline,NVIDIA B200-SXM-180GB,Tokens/s,3,2,11586.15,11732.125,12357.25,5.0-0056
+v5.0,llama2-70b-interactive-99,Offline,NVIDIA H100-NVL-94GB,Tokens/s,1,1,1944.1625,1944.1625,1944.1625,5.0-0023
+v5.0,llama2-70b-interactive-99,Offline,NVIDIA H100-SXM-80GB,Tokens/s,3,2,2923.6625,2929.425,3873.6875,5.0-0020
+v5.0,llama2-70b-interactive-99,Offline,NVIDIA H200-NVL-141GB,Tokens/s,1,1,3916.85,3916.85,3916.85,5.0-0017
+v5.0,llama2-70b-interactive-99,Offline,NVIDIA H200-SXM-141GB,Tokens/s,12,9,4317.3,4373.49375,4437.3125,5.0-0053
+v5.0,llama2-70b-interactive-99,Server,NVIDIA B200-SXM-180GB,Tokens/s,3,2,7426.1625,7438.1625,7783.2125,5.0-0074
+v5.0,llama2-70b-interactive-99,Server,NVIDIA H100-NVL-94GB,Tokens/s,1,1,1380.1875,1380.1875,1380.1875,5.0-0023
+v5.0,llama2-70b-interactive-99,Server,NVIDIA H100-SXM-80GB,Tokens/s,3,2,1622.425,1750.6125,1751.0625,5.0-0039
+v5.0,llama2-70b-interactive-99,Server,NVIDIA H200-NVL-141GB,Tokens/s,1,1,2016.8,2016.8,2016.8,5.0-0017
+v5.0,llama2-70b-interactive-99,Server,NVIDIA H200-SXM-141GB,Tokens/s,12,9,2422.625,2428.2875000000004,2529.3125,5.0-0033
+v5.0,llama2-70b-interactive-99.9,Offline,NVIDIA B200-SXM-180GB,Tokens/s,3,2,11586.15,11732.125,12357.25,5.0-0056
+v5.0,llama2-70b-interactive-99.9,Offline,NVIDIA H100-NVL-94GB,Tokens/s,1,1,1944.1625,1944.1625,1944.1625,5.0-0023
+v5.0,llama2-70b-interactive-99.9,Offline,NVIDIA H100-SXM-80GB,Tokens/s,3,2,2923.125,2956.7125,3873.6875,5.0-0020
+v5.0,llama2-70b-interactive-99.9,Offline,NVIDIA H200-NVL-141GB,Tokens/s,1,1,3916.85,3916.85,3916.85,5.0-0017
+v5.0,llama2-70b-interactive-99.9,Offline,NVIDIA H200-SXM-141GB,Tokens/s,11,8,4317.3,4373.3375,4437.3125,5.0-0053
+v5.0,llama2-70b-interactive-99.9,Server,NVIDIA B200-SXM-180GB,Tokens/s,3,2,7426.1625,7438.1625,7783.2125,5.0-0074
+v5.0,llama2-70b-interactive-99.9,Server,NVIDIA H100-NVL-94GB,Tokens/s,1,1,1380.1875,1380.1875,1380.1875,5.0-0023
+v5.0,llama2-70b-interactive-99.9,Server,NVIDIA H100-SXM-80GB,Tokens/s,3,2,1622.425,1748.45,1751.7625,5.0-0040
+v5.0,llama2-70b-interactive-99.9,Server,NVIDIA H200-NVL-141GB,Tokens/s,1,1,2016.8,2016.8,2016.8,5.0-0017
+v5.0,llama2-70b-interactive-99.9,Server,NVIDIA H200-SXM-141GB,Tokens/s,11,8,2422.625,2428.025,2529.3125,5.0-0033
+v5.0,llama3.1-405b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,4,3,187.32,190.50187499999998,192.27125,5.0-0072
+v5.0,llama3.1-405b,Offline,NVIDIA GB200,Tokens/s,2,2,192.86666666666667,196.67158333333333,200.4765,5.0-0076
+v5.0,llama3.1-405b,Offline,NVIDIA H100-SXM-80GB,Tokens/s,2,2,49.66375,52.0955,54.52725,5.0-0065
+v5.0,llama3.1-405b,Offline,NVIDIA H200-SXM-141GB,Tokens/s,7,7,68.902625,69.81575,71.696625,5.0-0011
+v5.0,llama3.1-405b,Server,NVIDIA B200-SXM-180GB,Tokens/s,4,3,105.7275,119.06275,135.03875,5.0-0074
+v5.0,llama3.1-405b,Server,NVIDIA GB200,Tokens/s,2,2,122.91805555555555,126.72402777777778,130.53,5.0-0076
+v5.0,llama3.1-405b,Server,NVIDIA H100-SXM-80GB,Tokens/s,2,2,32.7475,33.6940625,34.640625,5.0-0020
+v5.0,llama3.1-405b,Server,NVIDIA H200-SXM-141GB,Tokens/s,7,7,34.55375,36.397875,36.784375,5.0-0011
+v5.0,mixtral-8x7b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,4,3,15186.125,15855.125,16099.375,5.0-0072
+v5.0,mixtral-8x7b,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,8324.43,8345.73,8351.7,5.0-0038
+v5.0,mixtral-8x7b,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,2,2,7718.34,7755.625,7792.91,5.0-0066
+v5.0,mixtral-8x7b,Offline,NVIDIA H100-NVL-94GB,Tokens/s,2,2,4714.2,4794.7125,4875.225,5.0-0050
+v5.0,mixtral-8x7b,Offline,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,3960.825,3960.825,3960.825,5.0-0064
+v5.0,mixtral-8x7b,Offline,NVIDIA H100-SXM-80GB,Tokens/s,2,2,6381.0875,6485.81875,6590.55,5.0-0010
+v5.0,mixtral-8x7b,Offline,NVIDIA H200-NVL-141GB,Tokens/s,1,1,6525.7375,6525.7375,6525.7375,5.0-0043
+v5.0,mixtral-8x7b,Offline,NVIDIA H200-SXM-141GB,Tokens/s,11,8,7588.575,7798.275,7939.3375,5.0-0053
+v5.0,mixtral-8x7b,Offline,Virtualized NVIDIA H100-SXM-80GB,Tokens/s,1,1,6439.3125,6439.3125,6439.3125,5.0-0022
+v5.0,mixtral-8x7b,Server,NVIDIA B200-SXM-180GB,Tokens/s,4,3,15301.375,15987.875,16130.875,5.0-0072
+v5.0,mixtral-8x7b,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Tokens/s,3,2,7735.98,7785.4,8042.27,5.0-0059
+v5.0,mixtral-8x7b,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Tokens/s,2,2,7182.77,7477.26,7771.75,5.0-0066
+v5.0,mixtral-8x7b,Server,NVIDIA H100-NVL-94GB,Tokens/s,2,2,4741.8125,4787.76875,4833.725,5.0-0050
+v5.0,mixtral-8x7b,Server,NVIDIA H100-PCIe-80GB,Tokens/s,1,1,3466.125,3466.125,3466.125,5.0-0064
+v5.0,mixtral-8x7b,Server,NVIDIA H100-SXM-80GB,Tokens/s,2,2,6365.5875,6514.0,6662.4125,5.0-0010
+v5.0,mixtral-8x7b,Server,NVIDIA H200-NVL-141GB,Tokens/s,1,1,6265.4875,6265.4875,6265.4875,5.0-0043
+v5.0,mixtral-8x7b,Server,NVIDIA H200-SXM-141GB,Tokens/s,11,8,7131.8,7453.4125,7725.25,5.0-0060
+v5.0,mixtral-8x7b,Server,Virtualized NVIDIA H100-SXM-80GB,Tokens/s,1,1,6362.0,6362.0,6362.0,5.0-0022
+v5.0,resnet,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Samples/s,1,1,96157.6,96157.6,96157.6,5.0-0066
+v5.0,resnet,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,43960.55,43960.55,43960.55,5.0-0024
+v5.0,resnet,Offline,NVIDIA H100-NVL-94GB,Samples/s,4,3,60491.0,65331.5,69712.25,5.0-0050
+v5.0,resnet,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,56567.0,56567.0,56567.0,5.0-0064
+v5.0,resnet,Offline,NVIDIA H100-SXM-80GB,Samples/s,4,3,89107.375,89600.5625,89772.125,5.0-0039
+v5.0,resnet,Offline,NVIDIA H200-NVL-141GB,Samples/s,2,2,85092.25,85476.375,85860.5,5.0-0043
+v5.0,resnet,Offline,NVIDIA H200-SXM-141GB,Samples/s,11,9,93726.25,95675.375,96662.5,5.0-0051
+v5.0,resnet,Offline,NVIDIA L40S,Samples/s,2,2,41260.25,42237.6875,43215.125,5.0-0018
+v5.0,resnet,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Queries/s,1,1,72995.1,72995.1,72995.1,5.0-0066
+v5.0,resnet,Server,NVIDIA GeForce RTX 4090,Queries/s,1,1,36862.65,36862.65,36862.65,5.0-0024
+v5.0,resnet,Server,NVIDIA H100-NVL-94GB,Queries/s,4,3,54994.5,60003.5625,66306.25,5.0-0050
+v5.0,resnet,Server,NVIDIA H100-PCIe-80GB,Queries/s,1,1,47003.5,47003.5,47003.5,5.0-0064
+v5.0,resnet,Server,NVIDIA H100-SXM-80GB,Queries/s,4,3,73020.125,77523.5,81500.25,5.0-0010
+v5.0,resnet,Server,NVIDIA H200-NVL-141GB,Queries/s,2,2,67515.625,72519.5625,77523.5,5.0-0043
+v5.0,resnet,Server,NVIDIA H200-SXM-141GB,Queries/s,11,9,76007.75,80025.375,84527.375,5.0-0033
+v5.0,resnet,Server,NVIDIA L40S,Queries/s,2,2,42505.875,42756.1875,43006.5,5.0-0045
+v5.0,retinanet,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Samples/s,1,1,1925.5,1925.5,1925.5,5.0-0066
+v5.0,retinanet,Offline,NVIDIA H100-NVL-94GB,Samples/s,6,4,1326.715,1350.1325000000002,1388.275,5.0-0023
+v5.0,retinanet,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,1177.4025,1177.4025,1177.4025,5.0-0064
+v5.0,retinanet,Offline,NVIDIA H100-SXM-80GB,Samples/s,4,3,1794.55,1819.3375,1829.725,5.0-0039
+v5.0,retinanet,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,1407.975,1693.6875,1725.3625,5.0-0071
+v5.0,retinanet,Offline,NVIDIA H200-SXM-141GB,Samples/s,13,9,1798.9125,1848.15,1900.0375,5.0-0070
+v5.0,retinanet,Offline,NVIDIA L40S,Samples/s,2,2,788.14125,801.7049999999999,815.26875,5.0-0018
+v5.0,retinanet,Offline,Virtualized NVIDIA H100-SXM-80GB,Samples/s,1,1,1789.1625,1789.1625,1789.1625,5.0-0007
+v5.0,retinanet,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Queries/s,1,1,1730.68,1730.68,1730.68,5.0-0066
+v5.0,retinanet,Server,NVIDIA H100-NVL-94GB,Queries/s,6,4,1200.27,1261.6731249999998,1319.435,5.0-0050
+v5.0,retinanet,Server,NVIDIA H100-PCIe-80GB,Queries/s,1,1,1109.9975,1109.9975,1109.9975,5.0-0064
+v5.0,retinanet,Server,NVIDIA H100-SXM-80GB,Queries/s,4,3,1608.725,1718.55,1738.5875,5.0-0010
+v5.0,retinanet,Server,NVIDIA H200-NVL-141GB,Queries/s,3,3,1478.6375,1498.55,1498.6125,5.0-0004
+v5.0,retinanet,Server,NVIDIA H200-SXM-141GB,Queries/s,13,9,1613.6875,1698.6,1823.6625,5.0-0070
+v5.0,retinanet,Server,NVIDIA L40S,Queries/s,2,2,761.92625,761.930625,761.935,5.0-0045
+v5.0,retinanet,Server,Virtualized NVIDIA H100-SXM-80GB,Queries/s,1,1,1608.725,1608.725,1608.725,5.0-0007
+v5.0,rgat,Offline,NVIDIA H100-NVL-94GB,Samples/s,1,1,8789.125,8789.125,8789.125,5.0-0023
+v5.0,rgat,Offline,NVIDIA H200-SXM-141GB,Samples/s,7,5,51252.125,55111.625,56271.875,5.0-0053
+v5.0,stable-diffusion-xl,Offline,AMD Instinct MI325X 256GB HBM3E,Samples/s,1,1,2.1370375,2.1370375,2.1370375,5.0-0002
+v5.0,stable-diffusion-xl,Offline,NVIDIA B200-SXM-180GB,Samples/s,3,3,3.7882125,3.7931125,3.7969,5.0-0056
+v5.0,stable-diffusion-xl,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Samples/s,3,2,2.46365,2.51236,2.52,5.0-0059
+v5.0,stable-diffusion-xl,Offline,NVIDIA GH200 Grace Hopper Superchip 96GB,Samples/s,1,1,2.41908,2.41908,2.41908,5.0-0066
+v5.0,stable-diffusion-xl,Offline,NVIDIA H100-NVL-94GB,Samples/s,1,1,1.6140275,1.6140275,1.6140275,5.0-0050
+v5.0,stable-diffusion-xl,Offline,NVIDIA H100-PCIe-80GB,Samples/s,1,1,1.336995,1.336995,1.336995,5.0-0064
+v5.0,stable-diffusion-xl,Offline,NVIDIA H100-SXM-80GB,Samples/s,4,3,2.2486125,2.2789937499999997,2.296875,5.0-0010
+v5.0,stable-diffusion-xl,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,2.0039375,2.0470375,2.0975125,5.0-0071
+v5.0,stable-diffusion-xl,Offline,NVIDIA H200-SXM-141GB,Samples/s,10,6,2.3732125,2.3954250000000004,2.430625,5.0-0070
+v5.0,stable-diffusion-xl,Offline,NVIDIA L40S,Samples/s,3,3,0.719065,0.7657325,0.77710125,5.0-0018
+v5.0,stable-diffusion-xl,Offline,TPU Trillium,Samples/s,1,1,1.5198125,1.5198125,1.5198125,5.0-0036
+v5.0,stable-diffusion-xl,Server,AMD Instinct MI325X 256GB HBM3E,Queries/s,1,1,2.0219625,2.0219625,2.0219625,5.0-0002
+v5.0,stable-diffusion-xl,Server,NVIDIA B200-SXM-180GB,Queries/s,3,3,3.5555375,3.55605,3.61505,5.0-0074
+v5.0,stable-diffusion-xl,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Queries/s,3,2,2.13661,2.16789,2.27968,5.0-0059
+v5.0,stable-diffusion-xl,Server,NVIDIA GH200 Grace Hopper Superchip 96GB,Queries/s,1,1,1.93745,1.93745,1.93745,5.0-0066
+v5.0,stable-diffusion-xl,Server,NVIDIA H100-NVL-94GB,Queries/s,1,1,1.475575,1.475575,1.475575,5.0-0050
+v5.0,stable-diffusion-xl,Server,NVIDIA H100-PCIe-80GB,Queries/s,1,1,1.0550275,1.0550275,1.0550275,5.0-0064
+v5.0,stable-diffusion-xl,Server,NVIDIA H100-SXM-80GB,Queries/s,4,3,2.0045625,2.00465,2.2239625,5.0-0010
+v5.0,stable-diffusion-xl,Server,NVIDIA H200-NVL-141GB,Queries/s,3,3,1.851225,1.9314875,1.993225,5.0-0071
+v5.0,stable-diffusion-xl,Server,NVIDIA H200-SXM-141GB,Queries/s,10,6,2.067975,2.08965,2.2875375,5.0-0070
+v5.0,stable-diffusion-xl,Server,NVIDIA L40S,Queries/s,3,3,0.62590625,0.63431,0.7431325,5.0-0018
+v5.0,stable-diffusion-xl,Server,TPU Trillium,Queries/s,1,1,1.3699875,1.3699875,1.3699875,5.0-0036
+v5.1,deepseek-r1,Offline,NVIDIA B200-SXM-180GB,Tokens/s,5,4,3811.6625,3859.2375,4172.375,5.1-0069
+v5.1,deepseek-r1,Offline,NVIDIA GB200,Tokens/s,1,1,4023.777777777778,4023.777777777778,4023.777777777778,5.1-0071
+v5.1,deepseek-r1,Offline,NVIDIA GB300,Tokens/s,2,2,5842.486111111111,5924.180555555556,6005.875,5.1-0097
+v5.1,deepseek-r1,Server,NVIDIA B200-SXM-180GB,Tokens/s,5,4,1739.1625,1828.8375,2324.025,5.1-0069
+v5.1,deepseek-r1,Server,NVIDIA GB200,Tokens/s,1,1,2327.472222222222,2327.472222222222,2327.472222222222,5.1-0071
+v5.1,deepseek-r1,Server,NVIDIA GB300,Tokens/s,2,2,2818.1375,2862.7354166666664,2907.3333333333335,5.1-0072
+v5.1,dlrm-v2-99,Offline,NVIDIA GH200 Grace Hopper Superchip 144GB,Samples/s,1,1,87228.0,87228.0,87228.0,5.1-0053
+v5.1,dlrm-v2-99,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,65772.5,71381.7,72557.5,5.1-0005
+v5.1,dlrm-v2-99,Offline,NVIDIA H200-SXM-141GB,Samples/s,2,2,80390.25,80686.4375,80982.625,5.1-0079
+v5.1,dlrm-v2-99,Server,NVIDIA GH200 Grace Hopper Superchip 144GB,Queries/s,1,1,80515.0,80515.0,80515.0,5.1-0053
+v5.1,dlrm-v2-99,Server,NVIDIA H200-NVL-141GB,Queries/s,3,3,50507.5,56262.5,65021.1,5.1-0050
+v5.1,dlrm-v2-99,Server,NVIDIA H200-SXM-141GB,Queries/s,2,2,73146.0,73520.625,73895.25,5.1-0007
+v5.1,dlrm-v2-99.9,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,39765.0,43359.1,43528.125,5.1-0005
+v5.1,dlrm-v2-99.9,Offline,NVIDIA H200-SXM-141GB,Samples/s,2,2,48507.25,48586.75,48666.25,5.1-0079
+v5.1,dlrm-v2-99.9,Server,NVIDIA H200-NVL-141GB,Queries/s,3,3,38506.25,41319.375,41356.5,5.1-0050
+v5.1,dlrm-v2-99.9,Server,NVIDIA H200-SXM-141GB,Queries/s,2,2,46258.25,46258.375,46258.5,5.1-0079
+v5.1,llama2-70b-99,Interactive,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,1056.8925,1080.9724999999999,1105.0525,5.1-0001
+v5.1,llama2-70b-99,Interactive,AMD Instinct MI325X 256GB HBM3e,Tokens/s,6,6,1945.775,2257.3,2355.7625,5.1-0068
+v5.1,llama2-70b-99,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,6,6,6517.5875,7484.60625,7856.4,5.1-0046
+v5.1,llama2-70b-99,Interactive,NVIDIA GB200,Tokens/s,1,1,7436.4,7436.4,7436.4,5.1-0070
+v5.1,llama2-70b-99,Interactive,NVIDIA H200-NVL-141GB,Tokens/s,2,2,1677.435,1864.9175,2052.4,5.1-0023
+v5.1,llama2-70b-99,Interactive,NVIDIA H200-SXM-141GB,Tokens/s,5,5,2444.3375,2554.0125,2884.9875,5.1-0075
+v5.1,llama2-70b-99,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,5,4,3348.2625,3459.7625,3481.8,5.1-0066
+v5.1,llama2-70b-99,Offline,"AMD Instinct MI300X 192GB HBM3 (x32), AMD Instinct MI325X 256GB HBM3e (x16)",Tokens/s,1,1,3524.9375,3524.9375,3524.9375,5.1-0065
+v5.1,llama2-70b-99,Offline,"AMD Instinct MI325X 256GB HBM3E (x16), AMD Instinct MI300X 192GB HBM3 (x8)",Tokens/s,1,1,3839.9249999999997,3839.9249999999997,3839.9249999999997,5.1-0091
+v5.1,llama2-70b-99,Offline,AMD Instinct MI325X 256GB HBM3e,Tokens/s,10,9,4082.50625,4234.36875,4319.375,5.1-0003
+v5.1,llama2-70b-99,Offline,NVIDIA B200-SXM-180GB,Tokens/s,10,8,9807.9375,12690.1875,12863.625,5.1-0061
+v5.1,llama2-70b-99,Offline,NVIDIA GB200,Tokens/s,3,3,12420.875,12934.225,13015.4,5.1-0008
+v5.1,llama2-70b-99,Offline,NVIDIA H100-SXM-80GB,Tokens/s,2,1,3896.975,3899.721875,3902.46875,5.1-0016
+v5.1,llama2-70b-99,Offline,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3764.4,3810.875,3908.425,5.1-0023
+v5.1,llama2-70b-99,Offline,NVIDIA H200-SXM-141GB,Tokens/s,9,7,4305.925,4370.575,4414.5875,5.1-0026
+v5.1,llama2-70b-99,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,3275.725,3275.725,3275.725,5.1-0051
+v5.1,llama2-70b-99,Offline,Virtualized NVIDIA H200-SXM-141GB,Tokens/s,1,1,4287.7375,4287.7375,4287.7375,5.1-0029
+v5.1,llama2-70b-99,Offline,[MAXSUN] MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,752.345,752.345,752.345,5.1-0056
+v5.1,llama2-70b-99,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,5,4,2972.60625,3074.225,3107.5625,5.1-0066
+v5.1,llama2-70b-99,Server,"AMD Instinct MI300X 192GB HBM3 (x32), AMD Instinct MI325X 256GB HBM3e (x16)",Tokens/s,1,1,3189.0833333333335,3189.0833333333335,3189.0833333333335,5.1-0065
+v5.1,llama2-70b-99,Server,"AMD Instinct MI325X 256GB HBM3E (x16), AMD Instinct MI300X 192GB HBM3 (x8)",Tokens/s,1,1,3358.120833333333,3358.120833333333,3358.120833333333,5.1-0091
+v5.1,llama2-70b-99,Server,AMD Instinct MI325X 256GB HBM3e,Tokens/s,9,9,3586.9875,3955.775,4017.425,5.1-0045
+v5.1,llama2-70b-99,Server,NVIDIA B200-SXM-180GB,Tokens/s,10,8,12383.25,12396.931250000001,12701.375,5.1-0073
+v5.1,llama2-70b-99,Server,NVIDIA GB200,Tokens/s,3,3,11504.125,12303.975,12339.9,5.1-0070
+v5.1,llama2-70b-99,Server,NVIDIA H100-SXM-80GB,Tokens/s,2,1,3818.83125,3819.946875,3821.0625,5.1-0016
+v5.1,llama2-70b-99,Server,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3201.8375,3434.025,3633.7875,5.1-0023
+v5.1,llama2-70b-99,Server,NVIDIA H200-SXM-141GB,Tokens/s,8,6,4145.175,4153.9875,4274.225,5.1-0007
+v5.1,llama2-70b-99,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,3250.7375,3250.7375,3250.7375,5.1-0051
+v5.1,llama2-70b-99,Server,Virtualized NVIDIA H200-SXM-141GB,Tokens/s,1,1,4173.0875,4173.0875,4173.0875,5.1-0029
+v5.1,llama2-70b-99,Server,[MAXSUN] MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,527.5125,527.5125,527.5125,5.1-0056
+v5.1,llama2-70b-99.9,Interactive,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,1056.8925,1080.9724999999999,1105.0525,5.1-0001
+v5.1,llama2-70b-99.9,Interactive,AMD Instinct MI325X 256GB HBM3e,Tokens/s,6,6,1945.775,2257.3,2355.7625,5.1-0068
+v5.1,llama2-70b-99.9,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,8,6,6517.5875,7483.9125,7856.4,5.1-0046
+v5.1,llama2-70b-99.9,Interactive,NVIDIA GB200,Tokens/s,1,1,7436.4,7436.4,7436.4,5.1-0070
+v5.1,llama2-70b-99.9,Interactive,NVIDIA H200-NVL-141GB,Tokens/s,2,2,1678.2425,1865.32125,2052.4,5.1-0023
+v5.1,llama2-70b-99.9,Interactive,NVIDIA H200-SXM-141GB,Tokens/s,5,5,2444.3375,2544.6,2884.9875,5.1-0075
+v5.1,llama2-70b-99.9,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,5,4,3348.2625,3459.7625,3481.8,5.1-0066
+v5.1,llama2-70b-99.9,Offline,"AMD Instinct MI300X 192GB HBM3 (x32), AMD Instinct MI325X 256GB HBM3e (x16)",Tokens/s,1,1,3524.9375,3524.9375,3524.9375,5.1-0065
+v5.1,llama2-70b-99.9,Offline,"AMD Instinct MI325X 256GB HBM3E (x16), AMD Instinct MI300X 192GB HBM3 (x8)",Tokens/s,1,1,3839.9249999999997,3839.9249999999997,3839.9249999999997,5.1-0091
+v5.1,llama2-70b-99.9,Offline,AMD Instinct MI325X 256GB HBM3e,Tokens/s,10,9,4082.50625,4234.36875,4319.375,5.1-0003
+v5.1,llama2-70b-99.9,Offline,NVIDIA B200-SXM-180GB,Tokens/s,11,7,9807.9375,12690.875,12863.625,5.1-0061
+v5.1,llama2-70b-99.9,Offline,NVIDIA GB200,Tokens/s,2,2,12420.875,12677.55,12934.225,5.1-0070
+v5.1,llama2-70b-99.9,Offline,NVIDIA H100-SXM-80GB,Tokens/s,2,1,3896.975,3899.721875,3902.46875,5.1-0016
+v5.1,llama2-70b-99.9,Offline,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3791.775,3811.35,3908.425,5.1-0023
+v5.1,llama2-70b-99.9,Offline,NVIDIA H200-SXM-141GB,Tokens/s,9,7,4305.925,4351.5125,4414.5875,5.1-0026
+v5.1,llama2-70b-99.9,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,3275.6625,3275.6625,3275.6625,5.1-0051
+v5.1,llama2-70b-99.9,Offline,Virtualized NVIDIA H200-SXM-141GB,Tokens/s,1,1,4310.7125,4310.7125,4310.7125,5.1-0029
+v5.1,llama2-70b-99.9,Offline,[MAXSUN] MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,752.345,752.345,752.345,5.1-0056
+v5.1,llama2-70b-99.9,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,5,4,2972.60625,3074.225,3107.5625,5.1-0066
+v5.1,llama2-70b-99.9,Server,"AMD Instinct MI300X 192GB HBM3 (x32), AMD Instinct MI325X 256GB HBM3e (x16)",Tokens/s,1,1,3189.0833333333335,3189.0833333333335,3189.0833333333335,5.1-0065
+v5.1,llama2-70b-99.9,Server,"AMD Instinct MI325X 256GB HBM3E (x16), AMD Instinct MI300X 192GB HBM3 (x8)",Tokens/s,1,1,3358.120833333333,3358.120833333333,3358.120833333333,5.1-0091
+v5.1,llama2-70b-99.9,Server,AMD Instinct MI325X 256GB HBM3e,Tokens/s,9,9,3586.9875,3955.775,4017.425,5.1-0045
+v5.1,llama2-70b-99.9,Server,NVIDIA B200-SXM-180GB,Tokens/s,11,7,12383.25,12397.45,12701.375,5.1-0073
+v5.1,llama2-70b-99.9,Server,NVIDIA GB200,Tokens/s,2,2,12303.975,12321.9375,12339.9,5.1-0070
+v5.1,llama2-70b-99.9,Server,NVIDIA H100-SXM-80GB,Tokens/s,2,1,3818.83125,3819.946875,3821.0625,5.1-0016
+v5.1,llama2-70b-99.9,Server,NVIDIA H200-NVL-141GB,Tokens/s,3,3,3069.3375,3432.675,3633.7875,5.1-0023
+v5.1,llama2-70b-99.9,Server,NVIDIA H200-SXM-141GB,Tokens/s,8,6,3628.3625,4148.925,4253.675,5.1-0075
+v5.1,llama2-70b-99.9,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,3250.125,3250.125,3250.125,5.1-0051
+v5.1,llama2-70b-99.9,Server,Virtualized NVIDIA H200-SXM-141GB,Tokens/s,1,1,4171.325,4171.325,4171.325,5.1-0029
+v5.1,llama2-70b-99.9,Server,[MAXSUN] MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,527.5125,527.5125,527.5125,5.1-0056
+v5.1,llama3.1-405b,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,10,7,89.926125,93.92093750000001,96.3335,5.1-0073
+v5.1,llama3.1-405b,Interactive,NVIDIA GB200,Tokens/s,1,1,137.79194444444445,137.79194444444445,137.79194444444445,5.1-0071
+v5.1,llama3.1-405b,Interactive,NVIDIA H200-SXM-141GB,Tokens/s,1,1,25.347,25.347,25.347,5.1-0075
+v5.1,llama3.1-405b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,13,10,201.98125,206.00375,207.58875,5.1-0028
+v5.1,llama3.1-405b,Offline,NVIDIA GB200,Tokens/s,3,3,205.1986111111111,211.80825,213.955,5.1-0074
+v5.1,llama3.1-405b,Offline,NVIDIA GB300,Tokens/s,1,1,223.66805555555555,223.66805555555555,223.66805555555555,5.1-0072
+v5.1,llama3.1-405b,Offline,NVIDIA H100-SXM-80GB,Tokens/s,2,1,48.7990625,48.872375,48.9456875,5.1-0015
+v5.1,llama3.1-405b,Offline,NVIDIA H200-SXM-141GB,Tokens/s,3,3,67.875625,68.76325,69.093625,5.1-0075
+v5.1,llama3.1-405b,Offline,Virtualized NVIDIA H200-SXM-141GB,Tokens/s,1,1,68.407625,68.407625,68.407625,5.1-0029
+v5.1,llama3.1-405b,Server,NVIDIA B200-SXM-180GB,Tokens/s,13,10,155.0525,155.58625,159.94125,5.1-0073
+v5.1,llama3.1-405b,Server,NVIDIA GB200,Tokens/s,3,3,146.002,149.027,161.3097222222222,5.1-0071
+v5.1,llama3.1-405b,Server,NVIDIA GB300,Tokens/s,1,1,170.11666666666667,170.11666666666667,170.11666666666667,5.1-0072
+v5.1,llama3.1-405b,Server,NVIDIA H100-SXM-80GB,Tokens/s,2,1,35.2103125,35.21359375,35.216875,5.1-0016
+v5.1,llama3.1-405b,Server,NVIDIA H200-SXM-141GB,Tokens/s,3,3,36.5181875,36.92,36.940375,5.1-0079
+v5.1,llama3.1-405b,Server,Virtualized NVIDIA H200-SXM-141GB,Tokens/s,1,1,34.666375,34.666375,34.666375,5.1-0029
+v5.1,llama3.1-8b,Interactive,NVIDIA B200-SXM-180GB,Queries/s,4,4,12853.625,12863.125,15283.625,5.1-0086
+v5.1,llama3.1-8b,Interactive,NVIDIA H200-NVL-141GB,Queries/s,1,1,5486.025,5486.025,5486.025,5.1-0080
+v5.1,llama3.1-8b,Interactive,NVIDIA H200-SXM-141GB,Queries/s,2,2,6684.4,6724.549999999999,6764.7,5.1-0079
+v5.1,llama3.1-8b,Interactive,NVIDIA RTX PRO 6000 Blackwell Server Edition,Queries/s,1,1,4506.5375,4506.5375,4506.5375,5.1-0051
+v5.1,llama3.1-8b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,9,7,17094.0,17869.375,18370.0,5.1-0062
+v5.1,llama3.1-8b,Offline,NVIDIA H100-SXM-80GB,Tokens/s,1,1,5777.08,5777.08,5777.08,5.1-0083
+v5.1,llama3.1-8b,Offline,NVIDIA H200-NVL-141GB,Tokens/s,2,2,6295.2,6321.15,6347.1,5.1-0011
+v5.1,llama3.1-8b,Offline,NVIDIA H200-SXM-141GB,Tokens/s,3,3,7114.6625,7288.0375,8254.6,5.1-0049
+v5.1,llama3.1-8b,Offline,NVIDIA L40S,Tokens/s,1,1,1642.22,1642.22,1642.22,5.1-0082
+v5.1,llama3.1-8b,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,5855.225,5855.225,5855.225,5.1-0051
+v5.1,llama3.1-8b,Offline,[MAXSUN] MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,1618.0925,1618.0925,1618.0925,5.1-0056
+v5.1,llama3.1-8b,Server,NVIDIA B200-SXM-180GB,Tokens/s,9,7,14953.5,16082.5,16099.25,5.1-0028
+v5.1,llama3.1-8b,Server,NVIDIA H100-SXM-80GB,Tokens/s,1,1,5103.99,5103.99,5103.99,5.1-0083
+v5.1,llama3.1-8b,Server,NVIDIA H200-NVL-141GB,Tokens/s,2,2,4837.1125,5637.78125,6438.45,5.1-0080
+v5.1,llama3.1-8b,Server,NVIDIA H200-SXM-141GB,Tokens/s,3,3,5161.375,7246.3,8114.325,5.1-0049
+v5.1,llama3.1-8b,Server,NVIDIA L40S,Tokens/s,1,1,1207.14,1207.14,1207.14,5.1-0082
+v5.1,llama3.1-8b,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,5757.5,5757.5,5757.5,5.1-0051
+v5.1,llama3.1-8b,Server,[MAXSUN] MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,1337.1125,1337.1125,1337.1125,5.1-0056
+v5.1,mixtral-8x7b,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,6627.1625,6646.950000000001,6666.7375,5.1-0027
+v5.1,mixtral-8x7b,Offline,AMD Instinct MI325X 256GB HBM3e,Tokens/s,6,6,8495.55,8562.506249999999,8754.45,5.1-0096
+v5.1,mixtral-8x7b,Offline,NVIDIA H200-SXM-141GB,Tokens/s,2,2,7671.55,7717.55,7763.55,5.1-0049
+v5.1,mixtral-8x7b,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,4092.35,4092.35,4092.35,5.1-0051
+v5.1,mixtral-8x7b,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,5975.2375,5975.387500000001,5975.5375,5.1-0001
+v5.1,mixtral-8x7b,Server,AMD Instinct MI325X 256GB HBM3e,Tokens/s,6,6,6975.0375,7208.112499999999,7472.95,5.1-0002
+v5.1,mixtral-8x7b,Server,NVIDIA H200-SXM-141GB,Tokens/s,2,2,7499.1,7559.24375,7619.3875,5.1-0049
+v5.1,mixtral-8x7b,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,1,1,3766.7875,3766.7875,3766.7875,5.1-0051
+v5.1,retinanet,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,880.225,880.225,880.225,5.1-0032
+v5.1,retinanet,Offline,NVIDIA H200-NVL-141GB,Samples/s,4,4,1654.7475,1661.43125,1677.675,5.1-0023
+v5.1,retinanet,Offline,NVIDIA H200-SXM-141GB,Samples/s,7,6,1806.4625,1839.85,1874.575,5.1-0049
+v5.1,retinanet,Offline,NVIDIA L4-PCIe-24GB,Samples/s,2,1,226.3775,227.81650000000002,229.2555,5.1-0018
+v5.1,retinanet,Offline,NVIDIA L40S,Samples/s,2,2,813.7075,814.253125,814.79875,5.1-0012
+v5.1,retinanet,Server,NVIDIA GeForce RTX 4090,Queries/s,1,1,724.485,724.485,724.485,5.1-0032
+v5.1,retinanet,Server,NVIDIA H200-NVL-141GB,Queries/s,4,4,1575.1625,1625.18125,1650.35,5.1-0011
+v5.1,retinanet,Server,NVIDIA H200-SXM-141GB,Queries/s,7,6,1700.4125,1700.4125,1800.75,5.1-0007
+v5.1,retinanet,Server,NVIDIA L4-PCIe-24GB,Queries/s,2,1,203.49625,208.77687500000002,214.0575,5.1-0018
+v5.1,retinanet,Server,NVIDIA L40S,Queries/s,2,2,762.66375,762.6700000000001,762.67625,5.1-0012
+v5.1,rgat,Offline,NVIDIA B200-SXM-180GB,Samples/s,5,5,76793.0,78795.5,81403.75,5.1-0069
+v5.1,stable-diffusion-xl,Offline,AMD Instinct MI325X 256GB HBM3e,Samples/s,1,1,2.3232,2.3232,2.3232,5.1-0004
+v5.1,stable-diffusion-xl,Offline,NVIDIA B200-SXM-180GB,Samples/s,9,8,3.9473,3.99057,4.0711375,5.1-0060
+v5.1,stable-diffusion-xl,Offline,NVIDIA H100-SXM-80GB,Samples/s,1,1,2.2625875,2.2625875,2.2625875,5.1-0092
+v5.1,stable-diffusion-xl,Offline,NVIDIA H200-NVL-141GB,Samples/s,2,2,2.089935,2.11313625,2.1363375,5.1-0023
+v5.1,stable-diffusion-xl,Offline,NVIDIA H200-SXM-141GB,Samples/s,2,2,2.3903375,2.39448125,2.398625,5.1-0079
+v5.1,stable-diffusion-xl,Offline,NVIDIA L4-PCIe-24GB,Samples/s,1,1,0.251585,0.251585,0.251585,5.1-0019
+v5.1,stable-diffusion-xl,Offline,NVIDIA L40S,Samples/s,1,1,0.69005125,0.69005125,0.69005125,5.1-0022
+v5.1,stable-diffusion-xl,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Samples/s,1,1,1.384475,1.384475,1.384475,5.1-0051
+v5.1,stable-diffusion-xl,Offline,Virtualized NVIDIA H200-SXM-141GB,Samples/s,1,1,2.3301875,2.3301875,2.3301875,5.1-0029
+v5.1,stable-diffusion-xl,Server,AMD Instinct MI325X 256GB HBM3e,Queries/s,1,1,2.02555,2.02555,2.02555,5.1-0004
+v5.1,stable-diffusion-xl,Server,NVIDIA B200-SXM-180GB,Queries/s,9,8,3.47838,3.5579125,3.588025,5.1-0009
+v5.1,stable-diffusion-xl,Server,NVIDIA H100-SXM-80GB,Queries/s,1,1,2.2323125,2.2323125,2.2323125,5.1-0092
+v5.1,stable-diffusion-xl,Server,NVIDIA H200-NVL-141GB,Queries/s,2,2,1.9566175,1.9918774999999997,2.0271375,5.1-0023
+v5.1,stable-diffusion-xl,Server,NVIDIA H200-SXM-141GB,Queries/s,2,2,2.296475,2.2996937500000003,2.3029125,5.1-0079
+v5.1,stable-diffusion-xl,Server,NVIDIA L4-PCIe-24GB,Queries/s,1,1,0.220214,0.220214,0.220214,5.1-0019
+v5.1,stable-diffusion-xl,Server,NVIDIA L40S,Queries/s,1,1,0.7378375,0.7378375,0.7378375,5.1-0022
+v5.1,stable-diffusion-xl,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Queries/s,1,1,1.3603125,1.3603125,1.3603125,5.1-0051
+v5.1,stable-diffusion-xl,Server,Virtualized NVIDIA H200-SXM-141GB,Queries/s,1,1,2.2437875,2.2437875,2.2437875,5.1-0029
+v5.1,whisper,Offline,NVIDIA B200-SXM-180GB,Samples/s,14,9,5239.7875,5457.11875,5666.675,5.1-0069
+v5.1,whisper,Offline,NVIDIA GB200,Samples/s,1,1,5568.15,5568.15,5568.15,5.1-0077
+v5.1,whisper,Offline,NVIDIA GeForce RTX 4090,Samples/s,1,1,52.4765,52.4765,52.4765,5.1-0033
+v5.1,whisper,Offline,NVIDIA H100-NVL-94GB,Samples/s,1,1,2767.2999999999997,2767.2999999999997,2767.2999999999997,5.1-0021
+v5.1,whisper,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,3,3765.425,3959.2625,3962.775,5.1-0052
+v5.1,whisper,Offline,NVIDIA H200-SXM-141GB,Samples/s,4,4,4170.475,4256.4375,4306.3375,5.1-0049
+v5.1,whisper,Offline,NVIDIA L4-PCIe-24GB,Samples/s,1,1,349.2525,349.2525,349.2525,5.1-0019
+v5.1,whisper,Offline,NVIDIA L40S,Samples/s,1,1,1159.0,1159.0,1159.0,5.1-0022
+v6.0,deepseek-r1,Interactive,NVIDIA B300-SXM-270GB,Tokens/s,1,1,616.83125,616.83125,616.83125,6.0-0039
+v6.0,deepseek-r1,Interactive,NVIDIA GB200,Tokens/s,1,1,3337.75,3337.75,3337.75,6.0-0075
+v6.0,deepseek-r1,Interactive,NVIDIA GB300,Tokens/s,1,1,3481.027777777778,3481.027777777778,3481.027777777778,6.0-0078
+v6.0,deepseek-r1,Offline,NVIDIA B200-SXM-180GB,Tokens/s,3,3,7139.6125,7279.425,7322.7,6.0-0083
+v6.0,deepseek-r1,Offline,NVIDIA B300-SXM-270GB,Tokens/s,7,6,8285.9875,8657.9,8790.7125,6.0-0073
+v6.0,deepseek-r1,Offline,NVIDIA GB200,Tokens/s,3,3,4276.027777777777,6751.958333333333,7322.890625,6.0-0015
+v6.0,deepseek-r1,Offline,NVIDIA GB300,Tokens/s,5,3,8660.798611111111,9360.222222222223,9820.5,6.0-0017
+v6.0,deepseek-r1,Server,NVIDIA B200-SXM-180GB,Tokens/s,3,3,5118.6875,5128.4875,6461.6125,6.0-0083
+v6.0,deepseek-r1,Server,NVIDIA B300-SXM-270GB,Tokens/s,7,6,4429.4625,5813.825,7551.675,6.0-0084
+v6.0,deepseek-r1,Server,NVIDIA GB200,Tokens/s,3,3,2177.5138888888887,4668.138888888889,4867.671875,6.0-0015
+v6.0,deepseek-r1,Server,NVIDIA GB300,Tokens/s,5,3,5399.6875,5573.7125,8063.75,6.0-0082
+v6.0,dlrm-v3,Offline,NVIDIA B200-SXM-180GB,Samples/s,3,3,1305.9125,1318.425,1342.15,6.0-0085
+v6.0,dlrm-v3,Offline,NVIDIA GB200,Samples/s,1,1,1453.2916666666667,1453.2916666666667,1453.2916666666667,6.0-0075
+v6.0,dlrm-v3,Server,NVIDIA B200-SXM-180GB,Queries/s,3,3,1250.875,1250.875,1250.875,6.0-0040
+v6.0,dlrm-v3,Server,NVIDIA GB200,Queries/s,1,1,1388.8444444444444,1388.8444444444444,1388.8444444444444,6.0-0075
+v6.0,gpt-oss-120b,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,1,1,1644.4125,1644.4125,1644.4125,6.0-0083
+v6.0,gpt-oss-120b,Interactive,NVIDIA B300-SXM-270GB,Tokens/s,1,1,3250.725,3250.725,3250.725,6.0-0006
+v6.0,gpt-oss-120b,Interactive,NVIDIA GB200,Tokens/s,1,1,8679.569444444445,8679.569444444445,8679.569444444445,6.0-0075
+v6.0,gpt-oss-120b,Interactive,NVIDIA GB300,Tokens/s,1,1,9405.541666666666,9405.541666666666,9405.541666666666,6.0-0078
+v6.0,gpt-oss-120b,Offline,AMD Instinct MI355X 288GB HBM3e,Tokens/s,5,5,11424.4375,11480.825,11875.5,6.0-0002
+v6.0,gpt-oss-120b,Offline,AMD Instinct MI355X 288GB HBM3e (x94),Tokens/s,1,1,10968.829787234043,10968.829787234043,10968.829787234043,6.0-0003
+v6.0,gpt-oss-120b,Offline,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,210.2595,210.2595,210.2595,6.0-0059
+v6.0,gpt-oss-120b,Offline,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,384.225,384.225,384.225,6.0-0101
+v6.0,gpt-oss-120b,Offline,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,400.4775,400.4775,400.4775,6.0-0060
+v6.0,gpt-oss-120b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,2,2,10740.15,11186.99375,11633.8375,6.0-0091
+v6.0,gpt-oss-120b,Offline,NVIDIA B300-SXM-270GB,Tokens/s,6,6,12995.125,13552.3125,13937.0,6.0-0010
+v6.0,gpt-oss-120b,Offline,NVIDIA GB200,Tokens/s,2,2,12102.171875,12159.016493055555,12215.861111111111,6.0-0075
+v6.0,gpt-oss-120b,Offline,NVIDIA GB300,Tokens/s,5,4,14226.0625,14529.861111111111,15371.7,6.0-0079
+v6.0,gpt-oss-120b,Offline,NVIDIA H200-SXM-141GB,Tokens/s,1,1,3585.0,3585.0,3585.0,6.0-0092
+v6.0,gpt-oss-120b,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,3,1,1824.2625,1898.7375,1904.08,6.0-0045
+v6.0,gpt-oss-120b,Server,AMD Instinct MI355X 288GB HBM3e,Tokens/s,5,5,9419.3625,9825.4375,10267.0125,6.0-0002
+v6.0,gpt-oss-120b,Server,AMD Instinct MI355X 288GB HBM3e (x94),Tokens/s,1,1,9575.04255319149,9575.04255319149,9575.04255319149,6.0-0003
+v6.0,gpt-oss-120b,Server,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,113.04775,113.04775,113.04775,6.0-0059
+v6.0,gpt-oss-120b,Server,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,237.9185,237.9185,237.9185,6.0-0101
+v6.0,gpt-oss-120b,Server,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,221.0595,221.0595,221.0595,6.0-0060
+v6.0,gpt-oss-120b,Server,NVIDIA B200-SXM-180GB,Tokens/s,2,2,8948.5125,9939.51875,10930.525,6.0-0083
+v6.0,gpt-oss-120b,Server,NVIDIA B300-SXM-270GB,Tokens/s,6,6,12541.0,12583.9375,13831.875,6.0-0010
+v6.0,gpt-oss-120b,Server,NVIDIA GB200,Tokens/s,2,2,11617.96875,12053.553819444445,12489.138888888889,6.0-0075
+v6.0,gpt-oss-120b,Server,NVIDIA GB300,Tokens/s,5,4,13365.675,14945.671875,15232.916666666666,6.0-0081
+v6.0,gpt-oss-120b,Server,NVIDIA H200-SXM-141GB,Tokens/s,1,1,3012.9,3012.9,3012.9,6.0-0092
+v6.0,gpt-oss-120b,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,3,1,1671.8125,1773.7599999999998,1782.3625,6.0-0047
+v6.0,llama2-70b-99,Interactive,AMD Instinct MI350X 288GB HBM3e,Tokens/s,3,2,6452.075,6477.0625,6841.3125,6.0-0071
+v6.0,llama2-70b-99,Interactive,AMD Instinct MI355X 288GB HBM3e,Tokens/s,6,6,8826.35,8963.2625,9201.025,6.0-0002
+v6.0,llama2-70b-99,Interactive,AMD Instinct MI355X 288GB HBM3e (x87),Tokens/s,1,1,9028.988505747126,9028.988505747126,9028.988505747126,6.0-0001
+v6.0,llama2-70b-99,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,1,1,7705.5875,7705.5875,7705.5875,6.0-0048
+v6.0,llama2-70b-99,Interactive,NVIDIA B300-SXM-270GB,Tokens/s,2,2,8183.1,8511.81875,8840.5375,6.0-0024
+v6.0,llama2-70b-99,Interactive,NVIDIA GB200,Tokens/s,1,1,10484.097222222223,10484.097222222223,10484.097222222223,6.0-0075
+v6.0,llama2-70b-99,Interactive,NVIDIA GB300,Tokens/s,1,1,11307.333333333334,11307.333333333334,11307.333333333334,6.0-0078
+v6.0,llama2-70b-99,Interactive,NVIDIA H200-NVL-141GB,Tokens/s,1,1,2042.975,2042.975,2042.975,6.0-0021
+v6.0,llama2-70b-99,Interactive,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,2,1,779.765,781.296875,782.82875,6.0-0005
+v6.0,llama2-70b-99,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,3384.0625,3416.95,3449.8375,6.0-0069
+v6.0,llama2-70b-99,Offline,AMD Instinct MI300X 192GB HBM3 (x8) and AMD Instinct MI325X 256GB HBM3e (x8) and AMD Instinct MI355X 288GB HBM3e (x8),Tokens/s,1,1,6326.791666666667,6326.791666666667,6326.791666666667,6.0-0031
+v6.0,llama2-70b-99,Offline,AMD Instinct MI325X 256GB HBM3e,Tokens/s,1,1,4235.05,4235.05,4235.05,6.0-0070
+v6.0,llama2-70b-99,Offline,AMD Instinct MI350X 288GB HBM3e,Tokens/s,4,3,9867.0875,9934.6625,10059.7625,6.0-0100
+v6.0,llama2-70b-99,Offline,AMD Instinct MI355X 288GB HBM3e,Tokens/s,7,7,12191.8875,12502.375,12935.0,6.0-0002
+v6.0,llama2-70b-99,Offline,AMD Instinct MI355X 288GB HBM3e (Power Cap 1000 W),Tokens/s,1,1,10233.8125,10233.8125,10233.8125,6.0-0033
+v6.0,llama2-70b-99,Offline,AMD Instinct MI355X 288GB HBM3e (x87),Tokens/s,1,1,11978.275862068966,11978.275862068966,11978.275862068966,6.0-0001
+v6.0,llama2-70b-99,Offline,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,424.415,424.415,424.415,6.0-0059
+v6.0,llama2-70b-99,Offline,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,614.795,614.795,614.795,6.0-0101
+v6.0,llama2-70b-99,Offline,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,817.665,817.665,817.665,6.0-0060
+v6.0,llama2-70b-99,Offline,NVIDIA B200-SXM-180GB,Tokens/s,2,2,12699.375,12846.8125,12994.25,6.0-0048
+v6.0,llama2-70b-99,Offline,NVIDIA B300-SXM-270GB,Tokens/s,5,5,13583.625,13703.125,14466.25,6.0-0010
+v6.0,llama2-70b-99,Offline,NVIDIA GB200,Tokens/s,1,1,12334.083333333334,12334.083333333334,12334.083333333334,6.0-0075
+v6.0,llama2-70b-99,Offline,NVIDIA GB300,Tokens/s,1,1,15650.694444444445,15650.694444444445,15650.694444444445,6.0-0078
+v6.0,llama2-70b-99,Offline,NVIDIA H200-NVL-141GB,Tokens/s,1,1,4000.5,4000.5,4000.5,6.0-0021
+v6.0,llama2-70b-99,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,11,5,3340.475,3431.15,3738.55,6.0-0047
+v6.0,llama2-70b-99,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,3058.4625,3120.675,3182.8875,6.0-0069
+v6.0,llama2-70b-99,Server,AMD Instinct MI300X 192GB HBM3 (x8) and AMD Instinct MI325X 256GB HBM3e (x8) and AMD Instinct MI355X 288GB HBM3e (x8),Tokens/s,1,1,5896.708333333333,5896.708333333333,5896.708333333333,6.0-0031
+v6.0,llama2-70b-99,Server,AMD Instinct MI325X 256GB HBM3e,Tokens/s,1,1,3986.1125,3986.1125,3986.1125,6.0-0070
+v6.0,llama2-70b-99,Server,AMD Instinct MI350X 288GB HBM3e,Tokens/s,4,3,9549.1375,9570.525,9672.2375,6.0-0071
+v6.0,llama2-70b-99,Server,AMD Instinct MI355X 288GB HBM3e,Tokens/s,7,7,11659.05,12198.375,12535.25,6.0-0002
+v6.0,llama2-70b-99,Server,AMD Instinct MI355X 288GB HBM3e (Power Cap 1000 W),Tokens/s,1,1,9544.9375,9544.9375,9544.9375,6.0-0033
+v6.0,llama2-70b-99,Server,AMD Instinct MI355X 288GB HBM3e (x87),Tokens/s,1,1,11682.528735632184,11682.528735632184,11682.528735632184,6.0-0001
+v6.0,llama2-70b-99,Server,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,276.565,276.565,276.565,6.0-0059
+v6.0,llama2-70b-99,Server,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,424.6425,424.6425,424.6425,6.0-0101
+v6.0,llama2-70b-99,Server,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,549.875,549.875,549.875,6.0-0060
+v6.0,llama2-70b-99,Server,NVIDIA B200-SXM-180GB,Tokens/s,2,2,12412.6875,12683.53125,12954.375,6.0-0048
+v6.0,llama2-70b-99,Server,NVIDIA B300-SXM-270GB,Tokens/s,5,5,12773.25,13414.75,14313.0,6.0-0010
+v6.0,llama2-70b-99,Server,NVIDIA GB200,Tokens/s,1,1,11251.444444444445,11251.444444444445,11251.444444444445,6.0-0075
+v6.0,llama2-70b-99,Server,NVIDIA GB300,Tokens/s,1,1,12059.416666666666,12059.416666666666,12059.416666666666,6.0-0078
+v6.0,llama2-70b-99,Server,NVIDIA H200-NVL-141GB,Tokens/s,1,1,3635.6125,3635.6125,3635.6125,6.0-0021
+v6.0,llama2-70b-99,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,11,5,3190.5125,3389.225,3590.975,6.0-0047
+v6.0,llama2-70b-99.9,Interactive,AMD Instinct MI350X 288GB HBM3e,Tokens/s,1,1,6841.3125,6841.3125,6841.3125,6.0-0071
+v6.0,llama2-70b-99.9,Interactive,AMD Instinct MI355X 288GB HBM3e,Tokens/s,6,6,8826.35,8963.2625,9201.025,6.0-0002
+v6.0,llama2-70b-99.9,Interactive,AMD Instinct MI355X 288GB HBM3e (x87),Tokens/s,1,1,9028.988505747126,9028.988505747126,9028.988505747126,6.0-0001
+v6.0,llama2-70b-99.9,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,1,1,7662.55,7662.55,7662.55,6.0-0048
+v6.0,llama2-70b-99.9,Interactive,NVIDIA B300-SXM-270GB,Tokens/s,2,2,8183.1,8511.81875,8840.5375,6.0-0024
+v6.0,llama2-70b-99.9,Interactive,NVIDIA GB200,Tokens/s,1,1,10484.097222222223,10484.097222222223,10484.097222222223,6.0-0075
+v6.0,llama2-70b-99.9,Interactive,NVIDIA GB300,Tokens/s,1,1,11307.333333333334,11307.333333333334,11307.333333333334,6.0-0078
+v6.0,llama2-70b-99.9,Interactive,NVIDIA H200-NVL-141GB,Tokens/s,1,1,2042.975,2042.975,2042.975,6.0-0021
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,3384.0625,3416.95,3449.8375,6.0-0069
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI300X 192GB HBM3 (x8) and AMD Instinct MI325X 256GB HBM3e (x8) and AMD Instinct MI355X 288GB HBM3e (x8),Tokens/s,1,1,6326.791666666667,6326.791666666667,6326.791666666667,6.0-0031
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI325X 256GB HBM3e,Tokens/s,1,1,4235.05,4235.05,4235.05,6.0-0070
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI350X 288GB HBM3e,Tokens/s,1,1,9873.575,9873.575,9873.575,6.0-0071
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI355X 288GB HBM3e,Tokens/s,7,7,12191.8875,12502.375,12935.0,6.0-0002
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI355X 288GB HBM3e (Power Cap 1000 W),Tokens/s,1,1,10233.8125,10233.8125,10233.8125,6.0-0033
+v6.0,llama2-70b-99.9,Offline,AMD Instinct MI355X 288GB HBM3e (x87),Tokens/s,1,1,11978.275862068966,11978.275862068966,11978.275862068966,6.0-0001
+v6.0,llama2-70b-99.9,Offline,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,424.415,424.415,424.415,6.0-0059
+v6.0,llama2-70b-99.9,Offline,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,614.795,614.795,614.795,6.0-0101
+v6.0,llama2-70b-99.9,Offline,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,817.665,817.665,817.665,6.0-0060
+v6.0,llama2-70b-99.9,Offline,NVIDIA B200-SXM-180GB,Tokens/s,2,2,12690.75,12881.125,13071.5,6.0-0048
+v6.0,llama2-70b-99.9,Offline,NVIDIA B300-SXM-270GB,Tokens/s,3,3,13583.625,13667.875,14119.25,6.0-0073
+v6.0,llama2-70b-99.9,Offline,NVIDIA GB200,Tokens/s,1,1,12334.083333333334,12334.083333333334,12334.083333333334,6.0-0075
+v6.0,llama2-70b-99.9,Offline,NVIDIA GB300,Tokens/s,1,1,15650.694444444445,15650.694444444445,15650.694444444445,6.0-0078
+v6.0,llama2-70b-99.9,Offline,NVIDIA H200-NVL-141GB,Tokens/s,1,1,4000.5,4000.5,4000.5,6.0-0021
+v6.0,llama2-70b-99.9,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,7,3,3300.075,3379.35,3737.5125,6.0-0047
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI300X 192GB HBM3,Tokens/s,2,2,3058.4625,3120.675,3182.8875,6.0-0069
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI300X 192GB HBM3 (x8) and AMD Instinct MI325X 256GB HBM3e (x8) and AMD Instinct MI355X 288GB HBM3e (x8),Tokens/s,1,1,5896.708333333333,5896.708333333333,5896.708333333333,6.0-0031
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI325X 256GB HBM3e,Tokens/s,1,1,3986.1125,3986.1125,3986.1125,6.0-0070
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI350X 288GB HBM3e,Tokens/s,1,1,9672.2375,9672.2375,9672.2375,6.0-0071
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI355X 288GB HBM3e,Tokens/s,7,7,11659.05,12198.375,12535.25,6.0-0002
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI355X 288GB HBM3e (Power Cap 1000 W),Tokens/s,1,1,9544.9375,9544.9375,9544.9375,6.0-0033
+v6.0,llama2-70b-99.9,Server,AMD Instinct MI355X 288GB HBM3e (x87),Tokens/s,1,1,11682.528735632184,11682.528735632184,11682.528735632184,6.0-0001
+v6.0,llama2-70b-99.9,Server,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,276.565,276.565,276.565,6.0-0059
+v6.0,llama2-70b-99.9,Server,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,424.6425,424.6425,424.6425,6.0-0101
+v6.0,llama2-70b-99.9,Server,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,549.875,549.875,549.875,6.0-0060
+v6.0,llama2-70b-99.9,Server,NVIDIA B200-SXM-180GB,Tokens/s,2,2,12413.1,12683.2375,12953.375,6.0-0048
+v6.0,llama2-70b-99.9,Server,NVIDIA B300-SXM-270GB,Tokens/s,3,3,13077.875,13414.75,13549.0,6.0-0024
+v6.0,llama2-70b-99.9,Server,NVIDIA GB200,Tokens/s,1,1,11251.444444444445,11251.444444444445,11251.444444444445,6.0-0075
+v6.0,llama2-70b-99.9,Server,NVIDIA GB300,Tokens/s,1,1,12059.416666666666,12059.416666666666,12059.416666666666,6.0-0078
+v6.0,llama2-70b-99.9,Server,NVIDIA H200-NVL-141GB,Tokens/s,1,1,3635.6125,3635.6125,3635.6125,6.0-0021
+v6.0,llama2-70b-99.9,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,7,3,3143.15,3395.025,3589.6375,6.0-0047
+v6.0,llama3.1-405b,Interactive,NVIDIA B300-SXM-270GB,Tokens/s,1,1,95.581875,95.581875,95.581875,6.0-0039
+v6.0,llama3.1-405b,Interactive,NVIDIA GB200,Tokens/s,1,1,194.58472222222224,194.58472222222224,194.58472222222224,6.0-0075
+v6.0,llama3.1-405b,Interactive,NVIDIA GB300,Tokens/s,1,1,255.07500000000002,255.07500000000002,255.07500000000002,6.0-0078
+v6.0,llama3.1-405b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,1,1,168.78,168.78,168.78,6.0-0040
+v6.0,llama3.1-405b,Offline,NVIDIA B300-SXM-270GB,Tokens/s,4,4,243.48875,244.085625,246.38875,6.0-0010
+v6.0,llama3.1-405b,Offline,NVIDIA GB200,Tokens/s,1,1,214.74305555555554,214.74305555555554,214.74305555555554,6.0-0075
+v6.0,llama3.1-405b,Offline,NVIDIA GB300,Tokens/s,1,1,270.99583333333334,270.99583333333334,270.99583333333334,6.0-0078
+v6.0,llama3.1-405b,Server,NVIDIA B200-SXM-180GB,Tokens/s,1,1,122.95325,122.95325,122.95325,6.0-0040
+v6.0,llama3.1-405b,Server,NVIDIA B300-SXM-270GB,Tokens/s,4,4,167.56375,181.621875,185.55,6.0-0089
+v6.0,llama3.1-405b,Server,NVIDIA GB200,Tokens/s,1,1,196.30416666666667,196.30416666666667,196.30416666666667,6.0-0075
+v6.0,llama3.1-405b,Server,NVIDIA GB300,Tokens/s,1,1,258.71666666666664,258.71666666666664,258.71666666666664,6.0-0078
+v6.0,llama3.1-8b,Interactive,NVIDIA B200-SXM-180GB,Tokens/s,2,2,15767.875,15930.8125,16093.75,6.0-0062
+v6.0,llama3.1-8b,Interactive,NVIDIA B300-SXM-270GB,Tokens/s,1,1,16079.125,16079.125,16079.125,6.0-0039
+v6.0,llama3.1-8b,Interactive,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,7,2,4280.125,5229.8,5510.9125,6.0-0047
+v6.0,llama3.1-8b,Offline,Intel(R) Arc Pro(R) B50,Tokens/s,1,1,430.46,430.46,430.46,6.0-0058
+v6.0,llama3.1-8b,Offline,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,836.8575,836.8575,836.8575,6.0-0059
+v6.0,llama3.1-8b,Offline,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,1154.35,1154.35,1154.35,6.0-0101
+v6.0,llama3.1-8b,Offline,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,1690.575,1690.575,1690.575,6.0-0060
+v6.0,llama3.1-8b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,2,2,19078.125,19564.25,20050.375,6.0-0062
+v6.0,llama3.1-8b,Offline,NVIDIA B300-SXM-270GB,Tokens/s,3,3,18838.375,20679.0,20843.125,6.0-0006
+v6.0,llama3.1-8b,Offline,NVIDIA H200-NVL-141GB,Tokens/s,2,2,6054.15,6351.674999999999,6649.2,6.0-0021
+v6.0,llama3.1-8b,Offline,NVIDIA RTX PRO 4500 Blackwell,Tokens/s,2,1,2333.0,2336.875,2340.75,6.0-0043
+v6.0,llama3.1-8b,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,10,4,5816.425,6087.8,6247.6,6.0-0052
+v6.0,llama3.1-8b,Server,Intel(R) Arc Pro(R) B50,Tokens/s,1,1,296.7825,296.7825,296.7825,6.0-0058
+v6.0,llama3.1-8b,Server,Intel(R) Arc Pro(R) B60,Tokens/s,1,1,786.305,786.305,786.305,6.0-0059
+v6.0,llama3.1-8b,Server,Intel(R) Arc Pro(R) B70,Tokens/s,1,1,1042.5325,1042.5325,1042.5325,6.0-0101
+v6.0,llama3.1-8b,Server,MS-Intel Arc Pro B60 Dual 48G Turbo,Tokens/s,1,1,1573.4725,1573.4725,1573.4725,6.0-0060
+v6.0,llama3.1-8b,Server,NVIDIA B200-SXM-180GB,Tokens/s,2,2,16251.0,16329.875,16408.75,6.0-0048
+v6.0,llama3.1-8b,Server,NVIDIA B300-SXM-270GB,Tokens/s,3,3,17685.5,17695.625,18508.375,6.0-0006
+v6.0,llama3.1-8b,Server,NVIDIA H200-NVL-141GB,Tokens/s,2,2,5634.475,6033.6875,6432.9,6.0-0013
+v6.0,llama3.1-8b,Server,NVIDIA RTX PRO 4500 Blackwell,Tokens/s,2,1,2081.34,2175.88,2270.42,6.0-0050
+v6.0,llama3.1-8b,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,10,4,5626.075,5948.5125,6099.2875,6.0-0047
+v6.0,mixtral-8x7b,Offline,NVIDIA B200-SXM-180GB,Tokens/s,1,1,15808.5,15808.5,15808.5,6.0-0048
+v6.0,mixtral-8x7b,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,4,1,3707.075,4111.6825,4204.6625,6.0-0046
+v6.0,mixtral-8x7b,Server,NVIDIA B200-SXM-180GB,Tokens/s,1,1,15901.125,15901.125,15901.125,6.0-0048
+v6.0,mixtral-8x7b,Server,NVIDIA RTX PRO 6000 Blackwell Server Edition,Tokens/s,4,1,3461.525,3897.415,4018.875,6.0-0046
+v6.0,qwen3-vl-235b-a22b,Offline,NVIDIA B200-SXM-180GB,Samples/s,1,1,9.8796625,9.8796625,9.8796625,6.0-0094
+v6.0,qwen3-vl-235b-a22b,Offline,NVIDIA B300-SXM-270GB,Samples/s,1,1,9.7846875,9.7846875,9.7846875,6.0-0084
+v6.0,qwen3-vl-235b-a22b,Offline,NVIDIA GB200,Samples/s,2,2,10.6184,10.779975,10.94155,6.0-0074
+v6.0,qwen3-vl-235b-a22b,Offline,NVIDIA GB300,Samples/s,3,3,15.15385,15.238125,15.311075,6.0-0077
+v6.0,qwen3-vl-235b-a22b,Offline,NVIDIA H200-SXM-141GB,Samples/s,1,1,2.27725,2.27725,2.27725,6.0-0095
+v6.0,qwen3-vl-235b-a22b,Server,NVIDIA B200-SXM-180GB,Queries/s,1,1,8.483025,8.483025,8.483025,6.0-0094
+v6.0,qwen3-vl-235b-a22b,Server,NVIDIA B300-SXM-270GB,Queries/s,1,1,5.6439125,5.6439125,5.6439125,6.0-0084
+v6.0,qwen3-vl-235b-a22b,Server,NVIDIA GB200,Queries/s,2,2,8.890625,9.199225,9.507825,6.0-0074
+v6.0,qwen3-vl-235b-a22b,Server,NVIDIA GB300,Queries/s,3,3,10.40855,10.408775,10.78845,6.0-0080
+v6.0,qwen3-vl-235b-a22b,Server,NVIDIA H200-SXM-141GB,Queries/s,1,1,1.381225,1.381225,1.381225,6.0-0095
+v6.0,wan-2.2-t2v-a14b,Offline,NVIDIA B200-SXM-180GB,Samples/s,1,1,0.005805025,0.005805025,0.005805025,6.0-0072
+v6.0,wan-2.2-t2v-a14b,Offline,NVIDIA B300-SXM-270GB,Samples/s,3,3,0.0071640375,0.0072856125,0.007315425,6.0-0073
+v6.0,wan-2.2-t2v-a14b,Offline,NVIDIA GB200,Samples/s,1,1,0.006745425,0.006745425,0.006745425,6.0-0074
+v6.0,wan-2.2-t2v-a14b,Offline,NVIDIA GB300,Samples/s,1,1,0.0091555,0.0091555,0.0091555,6.0-0077
+v6.0,wan-2.2-t2v-a14b,Offline,NVIDIA H200-SXM-141GB,Samples/s,1,1,0.0022667125,0.0022667125,0.0022667125,6.0-0061
+v6.0,whisper,Offline,Intel(R) Arc Pro(R) B50,Samples/s,1,1,528.0875,528.0875,528.0875,6.0-0058
+v6.0,whisper,Offline,Intel(R) Arc Pro(R) B60,Samples/s,1,1,998.9075,998.9075,998.9075,6.0-0059
+v6.0,whisper,Offline,Intel(R) Arc Pro(R) B70,Samples/s,1,1,1399.15,1399.15,1399.15,6.0-0101
+v6.0,whisper,Offline,NVIDIA B200-SXM-180GB,Samples/s,5,5,5298.7125,5754.2375,6165.9,6.0-0040
+v6.0,whisper,Offline,NVIDIA B300-SXM-270GB,Samples/s,4,4,5707.35,5961.581249999999,6320.275,6.0-0098
+v6.0,whisper,Offline,NVIDIA GB200,Samples/s,1,1,5512.85,5512.85,5512.85,6.0-0041
+v6.0,whisper,Offline,NVIDIA H200-NVL-141GB,Samples/s,3,2,3774.4,3834.65,3996.15,6.0-0021
+v6.0,whisper,Offline,NVIDIA H200-SXM-141GB,Samples/s,1,1,4549.4625,4549.4625,4549.4625,6.0-0093
+v6.0,whisper,Offline,NVIDIA L40S,Samples/s,1,1,1823.455,1823.455,1823.455,6.0-0090
+v6.0,whisper,Offline,NVIDIA RTX PRO 6000 Blackwell Server Edition,Samples/s,11,6,1888.94,2292.2875,2356.33,6.0-0052

csv/claims.csv ADDED Viewed

The diff for this file is too large to render. See raw diff

csv/claims_meta.csv ADDED Viewed

	@@ -0,0 +1,17 @@

+key,value
+built,2026-10-28
+claims,531
+accelerator_rows,1727
+accelerator_groups,569
+training_v60.result_sets,229
+training_v60.commit,eabf23a07b2a
+training_v60.url,https://github.com/mlcommons/training_results_v6.0
+training_v50.result_sets,201
+training_v50.commit,d7e703fccc1c
+training_v50.url,https://github.com/mlcommons/training_results_v5.0
+training_v51.result_sets,185
+training_v51.commit,441d65f4445c
+training_v51.url,https://github.com/mlcommons/training_results_v5.1
+storage_v30.result_rows,144
+storage_v30.commit,802035fe6580
+storage_v30.url,https://github.com/mlcommons/storage_results_v3.0

csv/claims_mlcommons_files.csv ADDED Viewed

	@@ -0,0 +1,5 @@

+version,url,sha256,rows,excluded_no_accelerator_count,excluded_latency_units,closed_datacenter_rows_without_accelerator_count_excluded
+v5.0,https://raw.githubusercontent.com/mlcommons/inference_results_v5.0/main/summary_results.json,5a9edf18b4272a85eaac8e15a96680b1562dee73d2fdae95efdfcd29ca63e570,17454,49,6,
+v5.1,https://raw.githubusercontent.com/mlcommons/inference_results_v5.1/main/summary_results.json,68c19b46cf5a87f535beae81154d61c3d981fe850ae3d09af530c769b6e0cc68,1448,61,0,
+v6.0,https://raw.githubusercontent.com/mlcommons/inference_results_v6.0/main/summary_results.json,52fa813a27834e8de38eca9fd381688df1cf9cd95020d2bd9e951f6cc16384ae,520,24,7,
+v6.1,https://raw.githubusercontent.com/mlcommons/inference_results_v6.1/main/summary.json,f40d71a728a8c1e647d5fef9869ec1fcbb54550433efbc12e96f90b09a2dd561,492,,,0

csv/claims_sources.csv ADDED Viewed

	@@ -0,0 +1,75 @@

+key,name,title_label,vendor,title,url,round
+NVIDIA v5.1,NVIDIA v5.1,"NVIDIA, MLPerf Inference v5.1",NVIDIA,NVIDIA Technical Blog: NVIDIA Blackwell Ultra Sets New Inference Records in MLPerf Debut,https://developer.nvidia.com/blog/nvidia-blackwell-ultra-sets-new-inference-records-in-mlperf-debut,MLPerf Inference v5.1
+NVIDIA Training v5.0,NVIDIA Training v5.0,"NVIDIA, MLPerf Training v5.0",NVIDIA,NVIDIA Technical Blog: NVIDIA Blackwell Delivers up to 2.6x Higher Performance in MLPerf Training v5.0 (post dated 4 June 2025),https://developer.nvidia.com/blog/nvidia-blackwell-delivers-up-to-2-6x-higher-performance-in-mlperf-training-v5-0/,MLPerf Training v5.0
+NVIDIA Training v5.1,NVIDIA Training v5.1,"NVIDIA, MLPerf Training v5.1",NVIDIA,NVIDIA Technical Blog: NVIDIA Blackwell Architecture Sweeps MLPerf Training v5.1 Benchmarks (post dated 12 November 2025),https://developer.nvidia.com/blog/nvidia-blackwell-architecture-sweeps-mlperf-training-v5-1-benchmarks/,MLPerf Training v5.1
+NVIDIA Training v5.1 per dollar,NVIDIA Training v5.1 per dollar,"NVIDIA, MLPerf Training v5.1, per dollar",NVIDIA,NVIDIA Technical Blog: NVIDIA Blackwell Enables 3x Faster Training and Nearly 2x Training Performance Per Dollar than Previous-Gen Architecture (post dated 11 December 2025),https://developer.nvidia.com/blog/nvidia-blackwell-enables-3x-faster-training-and-nearly-2x-training-performance-per-dollar-than-previous-gen-architecture/,MLPerf Training v5.1
+Broadcom v5.1,Broadcom v5.1,"Broadcom, MLPerf Inference v5.1",Broadcom,Broadcom VMware Cloud Foundation Blog: The New Paradigm: MLPerf Inference 5.1 Confirms VCF is the Future of AI/ML Performance (post dated 15 December 2025),https://blogs.vmware.com/cloud-foundation/2025/12/15/mlperf-5-1-confirms-vcf-future-of-ai-ml-performance/,MLPerf Inference v5.1
+Google v5.0,Google v5.0,"Google, MLPerf Inference v5.0",Google,Google Cloud Blog: AI Hypercomputer inference updates for Google Cloud TPU and GPU (post dated April 2025),https://cloud.google.com/blog/products/compute/ai-hypercomputer-inference-updates-for-google-cloud-tpu-and-gpu,MLPerf Inference v5.0
+K-Veritas,K-Veritas,"K-Veritas, kveritas-go local-mode tests",K-Veritas,"K-Veritas README, kveritas-go commit 521070c (26 October 2026)",https://github.com/27-GROUP/kveritas-go,kveritas-go local-mode tests
+Nebius Storage v3.0,Nebius Storage v3.0,"Nebius, MLPerf Storage v3.0",Nebius,Nebius blog: MLPerf Storage v3.0 results: Leading RetinaNet training on Nebius Object Storage (2026),https://nebius.com/blog/posts/mlperf-storage-v3-0-results,MLPerf Storage v3.0
+Azure Lustre Storage v3.0,Azure Lustre Storage v3.0,"Azure, MLPerf Storage v3.0",Azure,Microsoft Learn: Azure Managed Lustre MLPerf Storage v3.0 results (2026),https://learn.microsoft.com/en-us/azure/azure-managed-lustre/mlperf-storage-v3-results,MLPerf Storage v3.0
+NVIDIA AIStore Storage v3.0,NVIDIA AIStore Storage v3.0,"NVIDIA, MLPerf Storage v3.0",NVIDIA,NVIDIA AIStore blog: NVIDIA AIStore at MLPerf Storage v3.0: Linear Scale-Out and Multi-Cloud Portability (2026),https://aistore.nvidia.com/blog/2026/09/01/mlperf-storage-v3,MLPerf Storage v3.0
+NVIDIA Training v6.0,NVIDIA Training v6.0,"NVIDIA, MLPerf Training v6.0",NVIDIA,NVIDIA Technical Blog: NVIDIA Blackwell Tops MLPerf Training 6.0 with Industry-Leading Scale and Performance,https://developer.nvidia.com/blog/nvidia-blackwell-tops-mlperf-training-6-0-with-industry-leading-scale-and-performance/,MLPerf Training v6.0
+AMD v5.1,AMD v5.1,"AMD, MLPerf Inference v5.1",AMD,ROCm Blogs: Technical Dive into AMD's MLPerf Inference v5.1 Submission,https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inference-v5.1/README.html,MLPerf Inference v5.1
+AMD v6.0,AMD v6.0,"AMD, MLPerf Inference v6.0",AMD,ROCm Blogs: AMD Instinct GPUs MLPerf Inference v6.0 Submission,https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inference-v6.0/README.html,MLPerf Inference v6.0
+Intel v5.1,Intel v5.1,"Intel, MLPerf Inference v5.1",Intel,Intel Newsroom: Intel Arc Pro B-Series GPUs and Xeon 6 Shine in MLPerf Inference v5.1,https://newsroom.intel.com/artificial-intelligence/intel-arc-pro-b-series-gpus-and-xeon-6-shine-in-mlperf-inference-v5-1,MLPerf Inference v5.1
+Intel v6.0,Intel v6.0,"Intel, MLPerf Inference v6.0",Intel,"Intel Newsroom: Intel Delivers Open, Scalable AI Performance in MLPerf Inference v6.0",https://newsroom.intel.com/artificial-intelligence/intel-delivers-ai-performance-mlperf-inference-v6-0,MLPerf Inference v6.0
+Lambda v6.0,Lambda v6.0,"Lambda, MLPerf Inference v6.0",Lambda,"Lambda: Lambda's MLPerf Inference v6.0: hardware leap, software maturity, research breakthrough",https://lambda.ai/blog/lambdas-mlperf-inference-v6.0-hardware-leap-software-maturity-research-breakthrough,MLPerf Inference v6.0
+Nebius v6.0,Nebius v6.0,"Nebius, MLPerf Inference v6.0",Nebius,"Nebius: MLPerf Inference v6.0 results, top-tier AI performance on NVIDIA Blackwell and Blackwell Ultra",https://nebius.com/blog/posts/mlperf-inference-v6-0-results,MLPerf Inference v6.0
+Supermicro v6.0,Supermicro v6.0,"Supermicro, MLPerf Inference v6.0",Supermicro,Supermicro Data Center Stories: Supermicro Leads Whisper Benchmark in MLPerf v6.0 with NVIDIA Blackwell Ultra GPUs,https://learn-more.supermicro.com/data-center-stories/supermicro-leads-whisper-benchmark-in-mlperf-v6-nvidia-b300-gpus,MLPerf Inference v6.0
+Supermicro Training v6.0,Supermicro Training v6.0,"Supermicro, MLPerf Training v6.0",Supermicro,Supermicro Data Center Stories: Supermicro NVIDIA Blackwell Systems Demonstrate Linear Scalability for MLPerf Training v6.0,https://learn-more.supermicro.com/data-center-stories/supermicro-nvidia-blackwell-systems-linear-scalability-mlperf-training,MLPerf Training v6.0
+Dell v6.1,Dell v6.1,"Dell, MLPerf Inference v6.1",Dell,"Dell Technologies: Faster Inference, Fewer Compromises: MLPerf v6.1 Results on the Dell PowerEdge R770 and R770AP (post dated 23 October 2026)",https://www.dell.com/en-us/infohub/articles/faster-inference-fewer-compromises-mlperf-v6-1-results-on-the-dell-poweredge-r770-and-r770ap,MLPerf Inference v6.1
+Dell MI355X v6.0,Dell MI355X v6.0,"Dell, MLPerf Inference v6.0, MI355X post",Dell,Dell Technologies Info Hub: MLPerf Inference v6.0: Dell Showcases Breakthrough Performance with AMD Instinct MI355X GPUs (post dated 16 April 2026),https://infohub.delltechnologies.com/en-us/p/mlperf-inference-v6-0-dell-showcases-breakthrough-performance-with-amd-instinct-tm-mi355x-gpus/,MLPerf Inference v6.0
+Dell Overview v6.1,Dell Overview v6.1,"Dell, MLPerf Inference v6.1, overview post",Dell,Dell Technologies Info Hub: Driving AI Innovation: Dell Technologies Advances AI Inference Performance with MLPerf Inference v6.1 (post dated 16 September 2026),https://infohub.delltechnologies.com/en-uk/p/driving-ai-innovation-dell-technologies-advances-ai-inference-performance-with-mlperf-inference-v6-1/,MLPerf Inference v6.1
+Dell YOLO v6.0,Dell YOLO v6.0,"Dell, MLPerf Inference v6.0, YOLO post",Dell,Dell Technologies Info Hub: Benchmark Builders to Performance Breakers: Dell's YOLO Journey in MLPerf Inference v6.0 (post dated 24 April 2026),https://infohub.delltechnologies.com/en-us/p/benchmark-builders-to-performance-breakers-dell-s-yolo-journey-in-mlperf-inference-v6-0/,MLPerf Inference v6.0
+HPE v6.0,HPE v6.0,"HPE, MLPerf Inference v6.0",HPE,HPE Community blog: HPE secures 18 top rankings in MLPerf Inference v6.0 AI benchmarks (post dated 1 April 2026),https://community.hpe.com/t5/ai-unlocked/hpe-secures-18-top-rankings-in-mlperf-inference-v6-0-ai/ba-p/7264016,MLPerf Inference v6.0
+AMD Training v6.0,AMD Training v6.0,"AMD, MLPerf Training v6.0",AMD,"AMD Blogs: AMD Instinct GPUs Deliver Breakthrough Performance, Multi-node Scale and Ecosystem Validation in MLPerf Training 6.0 (post dated 16 June 2026)",https://www.amd.com/en/blogs/2026/amd-delivers-breakthrough-mlperf-training-6-0-results.html,MLPerf Training v6.0
+Lambda Training v6.0,Lambda Training v6.0,"Lambda, MLPerf Training v6.0",Lambda,Lambda: MLPerf Training v6.0: Lambda delivers fastest LLM training on NVIDIA GB300 NVL72 and fastest MoE training on NVIDIA HGX B200 (post dated 16 June 2026),https://lambda.ai/blog/mlperf-training-v6.0-lambda-delivers-fastest-llm-training,MLPerf Training v6.0
+CoreWeave Training v6.0,CoreWeave Training v6.0,"CoreWeave, MLPerf Training v6.0",CoreWeave,"CoreWeave: Scaling frontier-class AI training on CoreWeave, a deep technical analysis of MLPerf Training v6.0 results (whitepaper page)",https://coreweave.com/resources/whitepapers/scaling-frontier-class-ai-training-on-coreweave-a-deep-technical-analysis-of-mlperf-r-training-v6-0-results,MLPerf Training v6.0
+NVIDIA v6.1,NVIDIA v6.1,"NVIDIA, MLPerf Inference v6.1",NVIDIA,NVIDIA Blog: NVIDIA Vera Rubin NVL72 Delivers Leading Performance in MLPerf Inference v6.1 Debut,https://blogs.nvidia.com/blog/vera-rubin-nvl72-mlperf-inference,MLPerf Inference v6.1
+AMD v6.1,AMD v6.1,"AMD, MLPerf Inference v6.1",AMD,AMD Blogs: AMD Delivers Its Broadest MLPerf Inference 6.1 Submission (post dated 16 September 2026),https://www.amd.com/en/blogs/2026/amd-delivers-its-broadest-mlperf-inference-6-1-submission.html,MLPerf Inference v6.1
+Crusoe v6.1,Crusoe v6.1,"Crusoe, MLPerf Inference v6.1",Crusoe,Crusoe: Serving 5.75 million tokens per second: Crusoe's MLPerf Inference v6.1 results on AMD MI355X (post dated 16 September 2026),https://www.crusoe.ai/resources/blog/serving-5-75-million-tokens-per-second-crusoes-mlperf-inference-v6-1-results-on-amd-mi355x,MLPerf Inference v6.1
+Lambda v5.1,Lambda v5.1,"Lambda, MLPerf Inference v5.1",Lambda,"Lambda: LLM performance up 15.4%: MLPerf v5.1 confirms NVIDIA HGX B200 on Lambda is built for enterprise inference (post dated 9 September 2025, updated 22 October 2026)",https://lambda.ai/blog/lambda-mlperf-inference-v5.1,MLPerf Inference v5.1
+Nebius v5.1,Nebius v5.1,"Nebius, MLPerf Inference v5.1",Nebius,Nebius: Nebius proves bare-metal-class performance for AI inference workloads in MLPerf Inference v5.1,https://nebius.com/blog/posts/bare-metal-class-performance-mlperf-inference,MLPerf Inference v5.1
+Lambda v6.1,Lambda v6.1,"Lambda, MLPerf Inference v6.1",Lambda,"Lambda: MLPerf Inference v6.1: pioneering agent, VLM benchmarks (post dated 16 September 2026, updated 22 October 2026)",https://lambda.ai/blog/mlperf-inference-v6.1,MLPerf Inference v6.1
+Lenovo v5.1,Lenovo v5.1,"Lenovo, MLPerf Inference v5.1",Lenovo,"Lenovo: Unleashing AI Excellence: Lenovo ThinkSystem Dominates MLPerf v5.1 with Exceptional Inference Performance (Lenovo Press LP2330, published 11 November 2025)",https://lenovopress.lenovo.com/lp2330-lenovo-thinksystem-dominates-mlperf-v51-with-exceptional-inference-performance,MLPerf Inference v5.1
+Nebius v6.1,Nebius v6.1,"Nebius, MLPerf Inference v6.1",Nebius,Nebius: MLPerf Inference v6.1 Results: NVIDIA Vera Rubin NVL72,https://nebius.com/blog/posts/mlperf-inference-v6-1-results,MLPerf Inference v6.1
+Red Hat v5.0,Red Hat v5.0,"Red Hat, MLPerf Inference v5.0",Red Hat,Red Hat: MLPerf Inference v5.0 results with Supermicro's GH200 Grace Hopper Superchip-based Server and Red Hat OpenShift (post dated 18 June 2025),https://www.redhat.com/en/blog/mlperf-inference-v50-results,MLPerf Inference v5.0
+Red Hat v6.0,Red Hat v6.0,"Red Hat, MLPerf Inference v6.0",Red Hat,"Red Hat: Red Hat AI tops MLPerf Inference v6.0 with vLLM on Qwen3-VL, Whisper, and GPT-OSS-120B (post dated 1 April 2026)",https://www.redhat.com/en/blog/red-hat-ai-tops-mlperf-inference-v60-vllm-qwen3-vl-whisper-and-gpt-oss-120b,MLPerf Inference v6.0
+CoreWeave v6.1,CoreWeave v6.1,"CoreWeave, MLPerf Inference v6.1",CoreWeave,CoreWeave: CoreWeave Leads Cloud Providers in MLPerf Inference v6.1 Performance with NVIDIA Blackwell Ultra,https://coreweave.com/blog/coreweave-leads-cloud-providers-in-mlperf-r-inference-v6-1-performance-with-nvidia-blackwell-ultra,MLPerf Inference v6.1
+Intel v6.1,Intel v6.1,"Intel, MLPerf Inference v6.1",Intel,Intel: Intel Software Optimizations Boost AI Inference in MLPerf v6.1,https://www.intel.com/content/www/us/en/newsroom/news/data-center/intel-software-optimizations-boost-ai-inference-in-mlperf-v6-1.html,MLPerf Inference v6.1
+Azure v6.1,Azure v6.1,"Azure, MLPerf Inference v6.1",Azure,"Azure: Azure Delivers Strong MLPerf Inference v6.1 Results with NVIDIA GB300 & GB200 (post dated 16 September 2026, read from a republished copy at argonsys.com)",https://argonsys.com/microsoft-cloud/library/azure-delivers-strong-mlperf-inference-v6-1-results-with-nvidia-gb300-gb200/,MLPerf Inference v6.1
+Cisco X210c v5.1,Cisco X210c v5.1,"Cisco, MLPerf Inference v5.1",Cisco,Cisco: AI Performance: MLPerf Inference on Cisco UCS X210c M8 Compute Node with Intel Xeon 6 Processor (white paper updated 2 October 2025),https://www.cisco.com/c/en/us/products/collateral/servers-unified-computing/ucs-c-series-rack-servers/mlperf-ucs-x210c-m8-compute-node-wp.html,MLPerf Inference v5.1
+Cisco C845A v6.1,Cisco C845A v6.1,"Cisco, MLPerf Inference v5.1 and v6.1",Cisco,Cisco: AI Performance: MLPerf Inference on Cisco UCS C845A M8 Rack Server with NVIDIA GPUs (white paper updated 5 October 2026),https://www.cisco.com/c/en/us/products/collateral/servers-unified-computing/ucs-c-series-rack-servers/mlperf-ucs-c845a-m8-rack-server-wp.html,MLPerf Inference v5.1 and v6.1
+HPE v5.1,HPE v5.1,"HPE, MLPerf Inference v5.1",HPE,HPE: HPE delivers several world records in latest MLPerf Inference benchmarks (post dated 9 September 2025),https://community.hpe.com/t5/ai-unlocked/hpe-delivers-several-world-records-in-latest-mlperf-inference/ba-p/7255374,MLPerf Inference v5.1
+Supermicro v5.0,Supermicro v5.0,"Supermicro, MLPerf Inference v5.0",Supermicro,"Supermicro: Industry's First-to-Market Supermicro NVIDIA HGX B200 Systems Demonstrate AI Performance Leadership on MLPerf Inference v5.0 Results (press release, 3 April 2025)",https://www.supermicro.com/en/pressreleases/first-to-market-supermicro-nvidia-hgx-b200-systems-performance-leadership-mlperf-results,MLPerf Inference v5.0
+Azure Training v6.0,Azure Training v6.0,"Azure, MLPerf Training v6.0",Azure,"Azure: Inside Llama 3.1 405B MLPerf Training on Azure: System-Level Insights at 8K+ GPU Scale (post dated 24 June 2026, read from a republished copy on the Azure Feeds listing)",,MLPerf Training v6.0
+AMD Repro v6.0,AMD Repro v6.0,"AMD, MLPerf Inference v6.0, reproduction post",AMD,AMD ROCm blog: Reproducing the AMD MLPerf Inference v6.0 Submission Result (post dated 1 April 2026),https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inf_v6.0-repro/README.html,MLPerf Inference v6.0
+Cisco Blog v6.1,Cisco Blog v6.1,"Cisco, MLPerf Inference v6.1, blog post",Cisco,Cisco Blogs: Industry's first MLPerf Inference benchmarking across multi-vendor accelerators (post dated 21 September 2026),https://blogs.cisco.com/datacenter/industrys-first-mlperf-inference-benchmarking-across-multi-vendor-accelerators/,MLPerf Inference v6.1
+AMD Tech v6.1,AMD Tech v6.1,"AMD, MLPerf Inference v6.1, technical dive",AMD,ROCm Blogs: Technical Dive into AMD MLPerf Inference v6.1 Submission (post dated 17 September 2026),https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inf-v6.1/README.html,MLPerf Inference v6.1
+AMD Repro v6.1,AMD Repro v6.1,"AMD, MLPerf Inference v6.1, reproduction post",AMD,ROCm Blogs: Reproducing AMD MLPerf Inference v6.1 Submission Results (post dated 17 September 2026),https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inf_v6.1-repro/README.html,MLPerf Inference v6.1
+AMD Repro v5.1,AMD Repro v5.1,"AMD, MLPerf Inference v5.1, reproduction post",AMD,ROCm Blogs: Reproducing the AMD Instinct GPUs MLPerf Inference v5.1 Submission (post dated 9 September 2025),https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inference5.1-repro/README.html,MLPerf Inference v5.1
+AMD Repro v5.0,AMD Repro v5.0,"AMD, MLPerf Inference v5.0",AMD,ROCm Blogs: Reproducing the AMD Instinct GPUs MLPerf Inference v5.0 Submission (post dated 2 April 2025),https://rocm.blogs.amd.com/artificial-intelligence/reproducing-amd-mlperf-inference-submission/README.html,MLPerf Inference v5.0
+AMD Pruning v5.1,AMD Pruning v5.1,"AMD, MLPerf Inference v5.1, pruning post",AMD,ROCm Blogs: Slim Down Your Llama: Pruning & Fine-Tuning for Maximum Performance (post dated 9 September 2025),https://rocm.blogs.amd.com/artificial-intelligence/mlperf-llama-pruning/README.html,MLPerf Inference v5.1
+NVIDIA v6.0,NVIDIA v6.0,"NVIDIA, MLPerf Inference v6.0",NVIDIA,NVIDIA Technical Blog: NVIDIA Platform Delivers Lowest Token Cost Enabled by Extreme Co-Design (post dated 1 April 2026),https://developer.nvidia.com/blog/?p=115040,MLPerf Inference v6.0
+NVIDIA v5.0,NVIDIA v5.0,"NVIDIA, MLPerf Inference v5.0",NVIDIA,NVIDIA Technical Blog: NVIDIA Blackwell Delivers Massive Performance Leaps in MLPerf Inference v5.0 (post dated 2 April 2025),https://developer.nvidia.com/blog/nvidia-blackwell-delivers-massive-performance-leaps-in-mlperf-inference-v5-0,MLPerf Inference v5.0
+Supp AMD v6.1,Supp AMD v6.1,"AMD, MLPerf Inference v6.1, MLCommons supplement",AMD,"MLCommons: MLPerf Inference v6.1 Results Discussion, AMD statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Crusoe v6.1,Supp Crusoe v6.1,"Crusoe, MLPerf Inference v6.1, MLCommons supplement",Crusoe,"MLCommons: MLPerf Inference v6.1 Results Discussion, Crusoe statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp ASUS v6.1,Supp ASUS v6.1,"ASUS, MLPerf Inference v6.1",ASUS,"MLCommons: MLPerf Inference v6.1 Results Discussion, ASUS statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Cisco v6.1,Supp Cisco v6.1,"Cisco, MLPerf Inference v6.1, MLCommons supplement",Cisco,"MLCommons: MLPerf Inference v6.1 Results Discussion, Cisco statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp CoreWeave v6.1,Supp CoreWeave v6.1,"CoreWeave, MLPerf Inference v6.1, MLCommons supplement",CoreWeave,"MLCommons: MLPerf Inference v6.1 Results Discussion, CoreWeave statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Fujitsu v6.1,Supp Fujitsu v6.1,"Fujitsu, MLPerf Inference v6.1",Fujitsu,"MLCommons: MLPerf Inference v6.1 Results Discussion, Fujitsu statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp HPE v6.1 v6.1,Supp HPE v6.1,"HPE, MLPerf Inference v6.1",HPE,"MLCommons: MLPerf Inference v6.1 Results Discussion, HPE statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Intel v6.1,Supp Intel v6.1,"Intel, MLPerf Inference v6.1, MLCommons supplement",Intel,"MLCommons: MLPerf Inference v6.1 Results Discussion, Intel statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp MangoBoost v6.1,Supp MangoBoost v6.1,"MangoBoost, MLPerf Inference v6.1",MangoBoost,"MLCommons: MLPerf Inference v6.1 Results Discussion, MangoBoost statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp NVIDIA v6.1,Supp NVIDIA v6.1,"NVIDIA, MLPerf Inference v6.1, MLCommons supplement",NVIDIA,"MLCommons: MLPerf Inference v6.1 Results Discussion, NVIDIA statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Lambda v6.1,Supp Lambda v6.1,"Lambda, MLPerf Inference v6.1, MLCommons supplement",Lambda,"MLCommons: MLPerf Inference v6.1 Results Discussion, Lambda statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp ScitiX v6.1,Supp ScitiX v6.1,"ScitiX, MLPerf Inference v6.1",ScitiX,"MLCommons: MLPerf Inference v6.1 Results Discussion, ScitiX statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp VibeHPC v6.1,Supp VibeHPC v6.1,"VibeHPC, MLPerf Inference v6.1",VibeHPC,"MLCommons: MLPerf Inference v6.1 Results Discussion, VibeHPC statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Wiwynn v6.1,Supp Wiwynn v6.1,"Wiwynn, MLPerf Inference v6.1",Wiwynn,"MLCommons: MLPerf Inference v6.1 Results Discussion, Wiwynn statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Orrick Industries v6.1,Supp Orrick Industries v6.1,"Orrick Industries, MLPerf Inference v6.1",Orrick Industries,"MLCommons: MLPerf Inference v6.1 Results Discussion, Orrick Industries statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Individual contributor v6.1,Supp Individual contributor v6.1,"Individual contributor Naeem Khoshnevis, MLPerf Inference v6.1",Individual contributor Naeem Khoshnevis,"MLCommons: MLPerf Inference v6.1 Results Discussion, Individual contributor statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Nebius v6.1,Supp Nebius v6.1,"Nebius, MLPerf Inference v6.1, MLCommons supplement",Nebius,"MLCommons: MLPerf Inference v6.1 Results Discussion, Nebius statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp Oracle v6.1,Supp Oracle v6.1,"Oracle, MLPerf Inference v6.1",Oracle,"MLCommons: MLPerf Inference v6.1 Results Discussion, Oracle statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Supp GigaComputing v6.1,Supp GigaComputing v6.1,"GigaComputing, MLPerf Inference v6.1",GigaComputing,"MLCommons: MLPerf Inference v6.1 Results Discussion, GigaComputing statement (supplement posted 16 September 2026)",https://mlcommons.org/wp-content/uploads/2026/09/Supplemental-MLCommons-MLPerf-Inference-v6.1.pdf,MLPerf Inference v6.1
+Apple Mac,Apple Mac,"Apple, MacBook Pro 14-inch with M5 Max, measured on our unit",Apple,"Apple: MacBook Pro Technical Specifications, 14-inch with M5 Max",https://www.apple.com/macbook-pro/specs/,"MacBook Pro 14-inch with M5 Max, measured on our unit"

csv/claims_tally.csv ADDED Viewed

	@@ -0,0 +1,75 @@

+source_key,ABOVE,BELOW,CONSISTENT,CANNOT_TELL
+NVIDIA v5.1,,,7,2
+NVIDIA Training v5.0,,1,4,1
+NVIDIA Training v5.1,,,7,2
+NVIDIA Training v5.1 per dollar,,,3,2
+Broadcom v5.1,1,2,2,3
+Google v5.0,1,,,4
+K-Veritas,,1,1,4
+Nebius Storage v3.0,,1,8,3
+Azure Lustre Storage v3.0,,,9,1
+NVIDIA AIStore Storage v3.0,,,15,2
+NVIDIA Training v6.0,,,13,2
+AMD v5.1,,,7,5
+AMD v6.0,,1,6,3
+Intel v5.1,,,,2
+Intel v6.0,1,1,1,2
+Lambda v6.0,,,5,3
+Nebius v6.0,,,7,
+Supermicro v6.0,,,2,1
+Supermicro Training v6.0,,3,1,
+Dell v6.1,,1,9,3
+Dell MI355X v6.0,,,5,2
+Dell Overview v6.1,,,9,1
+Dell YOLO v6.0,,,3,4
+HPE v6.0,,1,8,5
+AMD Training v6.0,,1,7,
+Lambda Training v6.0,,,6,2
+CoreWeave Training v6.0,,,3,
+NVIDIA v6.1,,1,4,3
+AMD v6.1,,,13,1
+Crusoe v6.1,,,3,2
+Lambda v5.1,1,,5,1
+Nebius v5.1,,,7,1
+Lambda v6.1,1,1,13,2
+Lenovo v5.1,,5,3,1
+Nebius v6.1,,,17,
+Red Hat v5.0,,,5,1
+Red Hat v6.0,2,1,13,4
+CoreWeave v6.1,,,11,
+Intel v6.1,,2,7,
+Azure v6.1,,1,6,
+Cisco X210c v5.1,,,3,
+Cisco C845A v6.1,,4,6,
+HPE v5.1,,3,10,3
+Supermicro v5.0,,2,4,1
+Azure Training v6.0,,,4,2
+AMD Repro v6.0,1,,4,1
+Cisco Blog v6.1,,,4,2
+AMD Tech v6.1,,4,13,2
+AMD Repro v6.1,,3,2,3
+AMD Repro v5.1,,2,6,
+AMD Repro v5.0,,2,3,1
+AMD Pruning v5.1,,,1,3
+NVIDIA v6.0,,1,8,1
+NVIDIA v5.0,,1,6,2
+Supp AMD v6.1,,,3,
+Supp Crusoe v6.1,,1,1,
+Supp ASUS v6.1,,,3,
+Supp Cisco v6.1,,,2,1
+Supp CoreWeave v6.1,,,1,
+Supp Fujitsu v6.1,,,,1
+Supp HPE v6.1 v6.1,,,4,
+Supp Intel v6.1,,1,2,1
+Supp MangoBoost v6.1,,,1,
+Supp NVIDIA v6.1,,1,4,1
+Supp Lambda v6.1,,,1,
+Supp ScitiX v6.1,,,1,
+Supp VibeHPC v6.1,,,4,
+Supp Wiwynn v6.1,,,1,1
+Supp Orrick Industries v6.1,,,2,
+Supp Individual contributor v6.1,,,1,
+Supp Nebius v6.1,,,2,
+Supp Oracle v6.1,,,1,
+Supp GigaComputing v6.1,,1,,
+Apple Mac,,1,5,3

csv/claims_vendor_page_hash.csv ADDED Viewed

	@@ -0,0 +1,75 @@

+source_key,sha256
+NVIDIA v5.1,3b067d2a407edbaa0af4c9b221fe3a2be8e651844b4738f6803b8c71e53affbd
+NVIDIA Training v5.0,687c9c325abc490755202e7e9bab43787ef2369a1fe66bfe5e0acd39ccc676be
+NVIDIA Training v5.1,3d1b7a2badc5547ca5b1194b02e4c654500f48e90405512d136253124c90c4c9
+NVIDIA Training v5.1 per dollar,315f59ce5b6540d2238574b49006a03710f61f3158f1ef45824878c0de034d13
+Broadcom v5.1,1f9be80f6bf9bc03d5bb34b93830153871980e6a75bc9d62f92a9d0bc8317250
+Google v5.0,f89264eaa5f67bf96076689b5f9f5df84a9320145be70a106c2c93789f7f484d
+K-Veritas,4552715d6d29cd8636579b3faf4a75d530841fb8032603b90dcf8329ea49563d
+Nebius Storage v3.0,c2ec8a6035c91b802f1adc558979f602fbcaa56555eb8f571599dd86e431a970
+Azure Lustre Storage v3.0,bf9e9c634dec23f57630d509c208212bf0f8752245d1b66570b7b2d57c948578
+NVIDIA AIStore Storage v3.0,0d21140d0423e6d99093d7f9bf720ba96c9fb9c132437c49316ca09b4af2b324
+NVIDIA Training v6.0,b9c095237fd10f8572de6a3158d805b057199557660f56a8a4d8f57844a1f8c0
+AMD v5.1,f702bbfb4cf37502c65183dc67491eb315236fbb0aed0e708bbb4a2db8a66e00
+AMD v6.0,463eb1911a2702c4eb5568a6642647fda4bd862788437c50048f2253cab8f920
+Intel v5.1,a7bccea373d09694bdc8fe39f81ff6f9c08f47a31e4152bef36d4089f468069e
+Intel v6.0,d6995b9fa777b04ab610b4bd70b7dc24b97cb02a6f1e39dcd9b056503014984d
+Lambda v6.0,499cfa555a35e65c190bf2b71e53fd427c1a1452a70cbe866e487c2e7c46fe45
+Nebius v6.0,9311dbdc61479772c0f4457488eb4e14dc9c68306ecdb8416eb4e80817f6cb63
+Supermicro v6.0,9f1f8b68b77916fe9fea574d62ae06598b37085d54d0e1e904f6c2cc26578252
+Supermicro Training v6.0,4b4641ba8793781ad77abcaed19f3b812b3bba615bc8305b961c819874e13146
+Dell v6.1,caf262f1011683d2338edd9f06d293afe56eb78a8369a5743e0e48260200be53
+Dell MI355X v6.0,ac69bf565eb2efdd237252c00791718cc3bd43c40e5feda9c04457847855f0d4
+Dell Overview v6.1,13842dfbe26e9b3c3bfd2c4687c1cd493ae13b8e748dfaf2cd61590a93954437
+Dell YOLO v6.0,6c16df203f8208938abec0b558126205c626b7e2cf4b425232834aa1a290a5df
+HPE v6.0,9f6d9caee586e99f80cf4816f25f47e40ae6c603ee3940bff943e8fc0b51658d
+AMD Training v6.0,8965c9a85416c6bce277d381d3e11510a210b12ccbd67ee2fafdf9a5c53552b8
+Lambda Training v6.0,f046b33097e7c72d848db6cb9b3ad4f09a2177644f85ba1ef84c2a1b270f3b0d
+CoreWeave Training v6.0,91148f0b1d8ce8444012830c422635acf0affca651710f6a18e3f9ffc6663a80
+NVIDIA v6.1,03e3d24d2cbf06c63dbd99e0cd2acba398d668e30320c4c71c4a0ef004c0b4f0
+AMD v6.1,b84e9f9a9a0d2432b61231a70126c593d822956cd17d5ceba8ee71eb41204673
+Crusoe v6.1,9e816d3dd3e30be5dff9b45062effaf529234cf05ff2b73f4de7220eba82ed4c
+Lambda v5.1,f3a4e2ab92bb8acaf4628ac3873ac885caad71942e881616e4dd2d7e80b7e235
+Nebius v5.1,7e155fb0d16adfd897a5c76ee07264707dc5be58707c82b7a4384829fccc62f5
+Lambda v6.1,d0dc8f380e301293942bf4446f23f1b10b6d85d088440ca31fecc9f3e4e42679
+Lenovo v5.1,99f11d790432fb7265661f8f249f8fd846db70a6217bd543cce84e8267891b78
+Nebius v6.1,273a1e81307ec6cb331dcbb46b9905ba7e79af1552ed6562a39aa224b0e8952a
+Red Hat v5.0,ade4137314eef68dc18848e243789496a7fc96a8db8bf62f6096cc957ea05db8
+Red Hat v6.0,d0c93ba732b73c92c991add3f95bb4f9f9bf9b67d0a8f8dbc6bcc7bc989e0e34
+CoreWeave v6.1,89b71107a98915e2b78a0e5222bef4c26c2452fb6e67215d6984d12a831e0dac
+Intel v6.1,5526357f1ff3c6c07b65069c599e908f56ba88dfac56fc90f3451cb8a99c66ef
+Azure v6.1,99add7010b670f9399e5f56ea49b1c4f05caac485a1b07afaecffb101516e25a
+Cisco X210c v5.1,7d39efb89cfb0bbf44af0d515645ae91f3fbf3042c20c9f15e64b36e4db66474
+Cisco C845A v6.1,971efba8adbef4ebb8796e766642d30bd7c479e48015023087dadf32506600c9
+HPE v5.1,d7e388ad4d53ae9c8afee15a9b53ded8dfb50bfb80e65a0b0e4b24a8871eb4e9
+Supermicro v5.0,e8f8e733f3ee3ee158f6d68c2952ebe268225f27c31595886b66437842da0e5a
+Azure Training v6.0,90b946c7fdd62af9a4e521f58179345099cd25e12b7a195dc25215f6ae3d0faf
+AMD Repro v6.0,ace123a48f5061512b78466771e8429d5c93753b6a1615c8fea4c4bf15023f41
+Cisco Blog v6.1,0d5c3f515762a315a1e665144ecd3ed5fd8438296f3e47cb5fcd8a3aa4e510e6
+AMD Tech v6.1,36ad96ad55318201114da9a3b27083cc47c005d0156fb22c69c0f5342a053397
+AMD Repro v6.1,9ff00526911e2781ecd1dd903a5609eb1b1ec94679a4dd258f171ed6db75a18d
+AMD Repro v5.1,9943931564be5bed1ab05b13e3768d9c7344df4c7b65ec5aac119f9b5c318162
+AMD Repro v5.0,1e834fad7e9cbd8b74fea708a253121704933bf23a91bf1defa3ce2d7710ce71
+AMD Pruning v5.1,b364e541aa0589de862f32181c399f67f04f20c39a15e2d9f1056330420cb452
+NVIDIA v6.0,3b3a1c660609fa981a69aef5959626b481885a2b680a5e6c43b4ff5b8d2749d2
+NVIDIA v5.0,e3b6ac94996fb3a123e511a2617f5be7983e3ec804c79a6fe686c583d068aa44
+Supp AMD v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Crusoe v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp ASUS v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Cisco v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp CoreWeave v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Fujitsu v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp HPE v6.1 v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Intel v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp MangoBoost v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp NVIDIA v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Lambda v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp ScitiX v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp VibeHPC v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Wiwynn v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Orrick Industries v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Individual contributor v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Nebius v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp Oracle v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Supp GigaComputing v6.1,a145f04ce685fa4b480dc5565e305b71b72322002ead5fa3883516fe275401fa
+Apple Mac,6a4ca4d93138eb835c22209922c38492005f885ef497a92805b7eb2a1d1bbed5

csv/mac_run.csv ADDED Viewed

	@@ -0,0 +1,18 @@

+kind,at,loadavg,powermode,thermal,threads,gib_per_array,reps,best_gbs,median_gbs,worst_gbs,mib_per_buffer
+env,2026-10-28T11:30:58Z,{ 3.72 4.13 4.30 },2,Note: No thermal warning level has been recorded Note: No performance warning level has been recorded Note: No CPU power status has been recorded ,,,,,,,
+cpu_triad,,,,,6,1.0,15,361.9,359.5,158.1,
+cpu_triad,,,,,6,1.0,15,362.9,361.1,158.9,
+cpu_triad,,,,,6,1.0,15,361.8,360.9,168.2,
+cpu_triad,,,,,12,1.0,15,302.1,280.6,187.8,
+cpu_triad,,,,,12,1.0,15,345.4,286.5,188.9,
+cpu_triad,,,,,12,1.0,15,287.0,278.7,191.2,
+cpu_triad,,,,,18,1.0,15,361.0,349.1,177.3,
+cpu_triad,,,,,18,1.0,15,358.2,350.4,206.1,
+cpu_triad,,,,,18,1.0,15,359.3,350.1,202.5,
+gpu_copyk,,,,,,,20,557.7,556.4,421.4,1024
+gpu_triad,,,,,,,20,560.1,559.4,538.0,1024
+gpu_copyk,,,,,,,20,557.9,556.9,533.3,1024
+gpu_triad,,,,,,,20,560.2,559.5,547.6,1024
+gpu_copyk,,,,,,,20,557.8,555.9,537.6,1024
+gpu_triad,,,,,,,20,560.2,558.9,545.0,1024
+env_after,2026-10-28T11:31:02Z,{ 3.50 4.08 4.28 },,,,,,,,,

csv/storage_v30_meta.csv ADDED Viewed

	@@ -0,0 +1,3 @@

+key,value
+source,https://github.com/mlcommons/storage_results_v3.0
+licence,Apache-2.0

csv/storage_v30_rows.csv ADDED Viewed

	@@ -0,0 +1,145 @@

+division,submitter,system,category,workload,runs,accelerators,hosts,au_pct,samples_per_s,io_MiB_per_s,io_GiB_per_s,uri_scheme,benchmark_API,product_API,storage_location,deployment,availability,submission_name,write_GiB_per_s,read_GiB_per_s,checkpoint_GB
+closed,Alluxio,aws_c5n18xlarge_i3en24xlarge_8c_10s_8a,training,unet3d,6,8,8,92.63041795978016,319.9516722797155,44732.20451913498,43.683793475717756,,file,file,remote,cloud,available,aws_c5n18xlarge_i3en24xlarge_8c_10s_8a,,,
+closed,Alluxio,aws_c7i48xlarge_i3en12xlarge_1c_1s_32a,training,retinanet,6,32,1,90.37935652358514,14564.284495896372,4485.738399449544,4.380603905712445,,file,file,remote,cloud,available,aws_c7i48xlarge_i3en12xlarge_1c_1s_32a,,,
+closed,Alluxio,aws_c7i48xlarge_i3en12xlarge_1c_1s_8a,training,retinanet,6,8,1,93.13657320344252,3748.975740110628,1154.6687680234033,1.1276062187728548,,file,file,remote,cloud,available,aws_c7i48xlarge_i3en12xlarge_1c_1s_8a,,,
+closed,Alluxio,aws_m7i48xlarge_i3en12xlarge_32c_32s_512p,checkpointing,llama3-405b,2,512,32,,,,,,file,file,remote,cloud,available,aws_m7i48xlarge_i3en12xlarge_32c_32s_512p,147.02534700759554,124.34398888798535,5292.248718261719
+closed,Alluxio,aws_r5n24xlarge_i3en12xlarge_1c_1s_8h,checkpointing,llama3-8b,2,8,1,,,,,,file,file,remote,cloud,available,aws_r5n24xlarge_i3en12xlarge_1c_1s_8h,4.643973844563218,5.67713748443572,104.70452117919922
+closed,Alluxio,aws_r5n24xlarge_i3en12xlarge_4c_10s_64p,checkpointing,llama3-70b,2,64,4,,,,,,file,file,remote,cloud,available,aws_r5n24xlarge_i3en12xlarge_4c_10s_64p,34.494582339203355,34.0776448290619,911.2121276855469
+closed,Alluxio,rigA_c5n18xl,training,unet3d,6,1,1,97.5410258394956,42.1183576653727,5888.536151954891,5.750523585893448,,file,file,remote,cloud,available,rigA_c5n18xl,,,
+closed,Azure,AMLFS-Durable-Premium-125-1280TiB-16-Standard_E192is_v6,checkpointing,llama3-70b,2,64,16,,,,,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-1280TiB-16-Standard_E192is_v6,105.50381161813687,123.59892086203688,911.2121276855469
+closed,Azure,AMLFS-Durable-Premium-125-160TiB-2-Standard_E192is_v6,checkpointing,llama3-8b,2,8,2,,,,,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-160TiB-2-Standard_E192is_v6,13.34064865889334,14.818591941034665,104.70452117919922
+closed,Azure,AMLFS-Durable-Premium-125-160TiB-3-Standard_D192ds_v6,training,unet3d,6,3,3,98.20510827238871,127.21469802722649,17785.79198991944,17.3689374901557,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-160TiB-3-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-125-2400TiB-46-Standard_D192ds_v6,training,unet3d,6,46,46,92.77719557885989,1842.725581446979,257630.08830241417,251.59188310782633,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-2400TiB-46-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5,checkpointing,llama3-1t,2,1024,128,,,,,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5,642.2033253302709,489.5958768747438,15426.589264154434
+closed,Azure,AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5,checkpointing,llama3-405b,2,512,128,,,,,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5,456.97121734565565,438.1095590967043,5292.248718261719
+closed,Azure,AMLFS-Durable-Premium-125-4096TiB-70-Standard_D192ds_v6,training,unet3d,6,70,70,91.8768355574808,2776.9432224078223,388242.35947163624,379.14292917151977,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-4096TiB-70-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-125-800TiB-16-Standard_D192ds_v6,training,unet3d,6,16,16,95.99397939512524,663.1905469956098,92720.17543146123,90.54704631978636,,file,file,remote,cloud,available,AMLFS-Durable-Premium-125-800TiB-16-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-20-25632TiB-70-Standard_D192ds_v6,training,unet3d,6,70,70,91.60451752198989,2768.7111586189276,387091.44077696075,378.01898513375073,,file,file,remote,cloud,available,AMLFS-Durable-Premium-20-25632TiB-70-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-20-4800TiB-16-Standard_D192ds_v6,training,unet3d,6,16,16,95.78464205233865,661.743771559797,92517.90283751945,90.34951448976508,,file,file,remote,cloud,available,AMLFS-Durable-Premium-20-4800TiB-16-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-20-960TiB-3-Standard_D192ds_v6,training,unet3d,6,3,3,98.17320004551571,127.17207287441481,17779.832599116304,17.363117772574515,,file,file,remote,cloud,available,AMLFS-Durable-Premium-20-960TiB-3-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-250-400TiB-16-Standard_D192ds_v6,training,unet3d,6,16,16,96.0973913226102,663.9028653332451,92819.76412663762,90.64430090491955,,file,file,remote,cloud,available,AMLFS-Durable-Premium-250-400TiB-16-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-250-80TiB-3-Standard_D192ds_v6,training,unet3d,6,3,3,98.225381172166,127.24153035136264,17789.543397131754,17.37260097376148,,file,file,remote,cloud,available,AMLFS-Durable-Premium-250-80TiB-3-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-40-2400TiB-16-Standard_D192ds_v6,training,unet3d,6,16,16,94.37566077762118,652.0094255355602,91156.95118468511,89.02046014129405,,file,file,remote,cloud,available,AMLFS-Durable-Premium-40-2400TiB-16-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-40-480TiB-3-Standard_D192ds_v6,training,unet3d,6,3,3,97.87151696783833,126.78031014244225,17725.060543934633,17.309629437436165,,file,file,remote,cloud,available,AMLFS-Durable-Premium-40-480TiB-3-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-500-200TiB-16-Standard_D192ds_v6,training,unet3d,6,16,16,95.68081570244276,661.0190848423504,92416.58492839227,90.25057121913308,,file,file,remote,cloud,available,AMLFS-Durable-Premium-500-200TiB-16-Standard_D192ds_v6,,,
+closed,Azure,AMLFS-Durable-Premium-500-40TiB-3-Standard_D192ds_v6,training,unet3d,6,3,3,98.19383172702176,127.19967213116847,17783.691230605502,17.366885967388185,,file,file,remote,cloud,available,AMLFS-Durable-Premium-500-40TiB-3-Standard_D192ds_v6,,,
+closed,Everpure,Everpure_FBEXA_file_32hosts_10-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,checkpointing,llama3-1t,1,1024,32,,,,,,file,file,remote,onprem,available,Everpure_FBEXA_file_32hosts_10-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,327.64719805437477,365.1207598586278,15426.589264154434
+closed,Everpure,Everpure_FBEXA_file_32hosts_15-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,checkpointing,llama3-1t,1,1024,32,,,,,,file,file,remote,onprem,available,Everpure_FBEXA_file_32hosts_15-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,484.09584386379964,439.87114303619046,15426.589264154434
+closed,Everpure,Everpure_FBEXA_file_32hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,checkpointing,llama3-1t,1,1024,32,,,,,,file,file,remote,onprem,available,Everpure_FBEXA_file_32hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,655.6004361087373,533.756143699845,15426.589264154434
+closed,Everpure,Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,checkpointing,llama3-1t,2,1024,32,,,,,,file,file,remote,onprem,available,Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,877.5237989244531,588.2768047121776,15426.589264154434
+closed,Everpure,Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,checkpointing,llama3-405b,1,512,32,,,,,,file,file,remote,onprem,available,Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,744.2387157606029,833.016210806981,5292.248718261719
+closed,Everpure,Everpure_FBEXA_file_51hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,kv_cache,llama3.1-8b,2,,,,,,,,file,file,remote,onprem,available,Everpure_FBEXA_file_51hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c,,,
+closed,FarmGPU,hickory,checkpointing,llama3-1t,1,1024,15,,,,,,file,file,remote,onprem,preview,hickory,169.67001937204188,252.1079035740201,15426.589264154434
+closed,FarmGPU,hickory,checkpointing,llama3-405b,1,512,15,,,,,,file,file,remote,onprem,preview,hickory,165.2149160928629,283.65453242888884,5292.248718261719
+closed,FarmGPU,hickory,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,preview,hickory,,,
+closed,FarmGPU,hickory,training,unet3d,6,75,15,91.96409502125466,2978.8917029423214,416476.6258195245,406.7154549018794,,file,file,remote,onprem,preview,hickory,,,
+closed,FarmGPU,potato1-qlc-closed-tcm,kv_cache,llama3.1-8b,1,,,,,,,,file,file,local,onprem,available,potato1-qlc-closed-tcm,,,
+closed,FarmGPU,potato8-qlc-closed-tcm,kv_cache,llama3.1-8b,1,,,,,,,,file,file,local,onprem,available,potato8-qlc-closed-tcm,,,
+closed,FarmGPU,potato8-qlc-np2-closed-tcm,kv_cache,llama3.1-8b,1,,,,,,,,file,file,local,onprem,available,potato8-qlc-np2-closed-tcm,,,
+closed,HPE,K3000-kvcache,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,preview,K3000-kvcache,,,
+closed,HPE,e2000,checkpointing,llama3-70b,2,64,16,,,,,,file,file,remote,onprem,available,e2000,48.18024515327689,110.22794945362989,911.2121276855469
+closed,HPE,e2000-unet3d,training,unet3d,6,30,15,94.55619792991766,1225.0496535146963,171273.27779430087,167.25906034599694,,file,file,remote,onprem,available,e2000-unet3d,,,
+closed,HPE,k3000,checkpointing,llama3-70b,1,64,8,,,,,,file,file,remote,onprem,preview,k3000,70.26223153852276,83.74542808097782,911.2121276855469
+closed,HPE,k3000-unet3d,training,unet3d,6,60,15,97.73403055373916,2532.6903360986644,354093.54572448286,345.7944782465653,,file,file,remote,onprem,preview,k3000-unet3d,,,
+closed,HolmesAI,StorMeshAI,training,retinanet,6,20,1,85.61695963222215,8623.848804413834,2656.1091788550175,2.593856619975603,,file,file,remote,cloud,available,StorMeshAI,,,
+closed,HolmesAI,StorMeshAI,training,unet3d,6,4,1,90.8331384551441,156.88409272415188,21933.84792000856,21.41977335938336,,file,file,remote,cloud,available,StorMeshAI,,,
+closed,InspurData,InspurData_InDataOS_AS13000G7_10clients,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_10clients,,,
+closed,InspurData,InspurData_InDataOS_AS13000G7_10clients,training,unet3d,6,50,10,91.20689736028054,1969.219883543951,275315.1622749616,268.8624631591422,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_10clients,,,
+closed,InspurData,InspurData_InDataOS_AS13000G7_10clients_checkpointing,checkpointing,llama3-70b,2,64,10,,,,,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_10clients_checkpointing,195.86855929267355,269.9591183910151,911.2121276855469
+closed,InspurData,InspurData_InDataOS_AS13000G7_1client,checkpointing,llama3-8b,2,8,1,,,,,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_1client,33.031401803633365,45.011544959951635,104.70452117919922
+closed,InspurData,InspurData_InDataOS_AS13000G7_KH-50000_10clients,checkpointing,llama3-70b,2,64,10,,,,,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_KH-50000_10clients,202.40849606043844,368.79434917650207,911.2121276855469
+closed,InspurData,InspurData_InDataOS_AS13000G7_KH-50000_10clients,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_KH-50000_10clients,,,
+closed,InspurData,InspurData_InDataOS_AS13000G7_KH-50000_10clients_Unet3D,training,unet3d,6,70,10,90.6251395681733,2739.234101116296,382970.27536646323,373.99440953756175,,file,file,remote,onprem,available,InspurData_InDataOS_AS13000G7_KH-50000_10clients_Unet3D,,,
+closed,NVIDIA,aistore-12n-oci-bm-denseio-e5-128-16client,checkpointing,llama3-1t,2,1024,16,,,,,,object,object,remote,cloud,available,aistore-12n-oci-bm-denseio-e5-128-16client,133.26620428586173,136.54350674628003,15426.589264154434
+closed,NVIDIA,aistore-12n-oci-bm-denseio-e5-128-16client,checkpointing,llama3-405b,2,512,16,,,,,,object,object,remote,cloud,available,aistore-12n-oci-bm-denseio-e5-128-16client,114.53121018453218,123.2881347348884,5292.248718261719
+closed,NVIDIA,aistore-12n-oci-bm-denseio-e5-128-1client,checkpointing,llama3-8b,2,8,1,,,,,,object,object,remote,cloud,available,aistore-12n-oci-bm-denseio-e5-128-1client,3.196488777385128,8.334717693512255,104.70452117919922
+closed,NVIDIA,aistore-12n-oci-bm-denseio-e5-128-20client-unet3d,training,unet3d,6,20,20,98.01536666413432,846.5426709017167,118354.49903773211,115.58056546653526,,object,object,remote,cloud,available,aistore-12n-oci-bm-denseio-e5-128-20client-unet3d,,,
+closed,NVIDIA,aistore-12n-oci-bm-denseio-e5-128-8client,checkpointing,llama3-70b,2,64,8,,,,,,object,object,remote,cloud,available,aistore-12n-oci-bm-denseio-e5-128-8client,27.724210916654336,78.12166947859546,911.2121276855469
+closed,NVIDIA,aistore-3n-aws-i8ge-150g-4client-m8gn24xl-unet3d,training,unet3d,6,8,4,98.37396216523095,339.92408507480917,47524.5326464581,46.41067641255674,,object,object,remote,cloud,available,aistore-3n-aws-i8ge-150g-4client-m8gn24xl-unet3d,,,
+closed,NVIDIA,aistore-3n-aws-i8ge-150g-5client-m8gn24xl,training,retinanet,6,255,5,88.76981905247644,114053.24730387353,35127.920712964136,34.30461007125404,,object,object,remote,cloud,available,aistore-3n-aws-i8ge-150g-5client-m8gn24xl,,,
+closed,NVIDIA,aistore-3n-gcp-z3-200g-8client-n2d96,training,unet3d,6,8,8,97.88103113179451,338.0440734228716,47261.68962046727,46.153993769987565,,object,object,remote,cloud,available,aistore-3n-gcp-z3-200g-8client-n2d96,,,
+closed,NVIDIA,aistore-3n-oci-bm-denseio-e5-128-1client,checkpointing,llama3-8b,2,8,1,,,,,,object,object,remote,cloud,available,aistore-3n-oci-bm-denseio-e5-128-1client,3.0202352379882162,6.992930457598599,104.70452117919922
+closed,NVIDIA,aistore-3n-oci-bm-denseio-e5-128-5client-retinanet,training,retinanet,6,190,5,94.06365488499786,90053.51900612167,27736.10528722767,27.086040319558272,,object,object,remote,cloud,available,aistore-3n-oci-bm-denseio-e5-128-5client-retinanet,,,
+closed,NVIDIA,aistore-3n-oci-bm-denseio-e5-128-5client-unet3d,training,unet3d,6,5,5,98.86264187253634,213.4684552990488,29844.865422277908,29.14537638894327,,object,object,remote,cloud,available,aistore-3n-oci-bm-denseio-e5-128-5client-unet3d,,,
+closed,NVIDIA,aistore-3n-oci-bm-denseio-e5-128-8client,checkpointing,llama3-70b,2,64,8,,,,,,object,object,remote,cloud,available,aistore-3n-oci-bm-denseio-e5-128-8client,22.14796788267677,33.28990554938302,911.2121276855469
+closed,NVIDIA,aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t,checkpointing,llama3-1t,2,1024,8,,,,,,object,object,remote,cloud,available,aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t,33.61921399099397,34.19594857148875,15426.589264154434
+closed,NVIDIA,aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t,checkpointing,llama3-405b,2,512,8,,,,,,object,object,remote,cloud,available,aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t,33.30672578549643,33.076730530779834,5292.248718261719
+closed,NVIDIA,aistore-6n-oci-bm-denseio-e5-128-10client-retinanet,training,retinanet,6,370,10,89.25373008545304,166384.46279087468,51245.715093185914,50.04464364568937,,object,object,remote,cloud,available,aistore-6n-oci-bm-denseio-e5-128-10client-retinanet,,,
+closed,NVIDIA,aistore-6n-oci-bm-denseio-e5-128-10client-unet3d,training,unet3d,6,10,10,98.74707647368561,426.4364823599876,59619.76634605895,58.22242807232319,,object,object,remote,cloud,available,aistore-6n-oci-bm-denseio-e5-128-10client-unet3d,,,
+closed,NVIDIA,aistore-6n-oci-bm-denseio-e5-128-1client,checkpointing,llama3-8b,2,8,1,,,,,,object,object,remote,cloud,available,aistore-6n-oci-bm-denseio-e5-128-1client,3.3973937004064005,11.08184250902501,104.70452117919922
+closed,NVIDIA,aistore-6n-oci-bm-denseio-e5-128-8client,checkpointing,llama3-70b,2,64,8,,,,,,object,object,remote,cloud,available,aistore-6n-oci-bm-denseio-e5-128-8client,26.541383802027422,56.79712068487546,911.2121276855469
+closed,NVIDIA,aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t,checkpointing,llama3-1t,2,1024,8,,,,,,object,object,remote,cloud,available,aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t,64.50496710106077,70.00709074663709,15426.589264154434
+closed,NVIDIA,aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t,checkpointing,llama3-405b,2,512,8,,,,,,object,object,remote,cloud,available,aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t,62.027056682203764,63.80655689735703,5292.248718261719
+closed,Nebius,nebius-s3-ckpt-llama3-70b,checkpointing,llama3-70b,1,64,8,,,,,s3,object,object,remote,cloud,available,nebius-s3-ckpt-llama3-70b,19.45538992050427,38.274839274839515,911.2121276855469
+closed,Nebius,nebius-s3-ckpt-llama3-8b,checkpointing,llama3-8b,1,8,1,,,,,s3,object,object,remote,cloud,available,nebius-s3-ckpt-llama3-8b,2.8118349276291923,7.143715604587631,104.70452117919922
+closed,Nebius,nebius-s3-retinanet-32nx24-cm64,training,retinanet,6,768,32,88.42844382305479,341881.3336218671,105298.0135560201,102.83009136330088,s3,object,object,remote,cloud,available,nebius-s3-retinanet-32nx24-cm64,,,
+closed,Nebius,nebius-s3-unet3d-21n,training,unet3d,6,21,21,92.47475811918585,838.2511080278598,117195.26182039271,114.44849787147726,s3,object,object,remote,cloud,available,nebius-s3-unet3d-21n,,,
+closed,NewFW,Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC,checkpointing,llama3-8b,2,8,1,,,,,,file,block,local,onprem,available,Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC,3.076140817854296,4.983509448925309,104.70452117919922
+closed,NewFW,Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC,kv_cache,llama3.1-8b,1,,,,,,,,file,block,local,onprem,available,Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC,,,
+closed,NewFW,Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC,training,unet3d,6,1,1,97.74196560717776,42.2128683477068,5901.749619918002,5.763427363201174,,file,block,local,onprem,available,Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC,,,
+closed,NewFW,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,checkpointing,llama3-8b,2,8,1,,,,,,file,block,local,onprem,available,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,8.63733219350259,12.999883786501638,104.70452117919922
+closed,NewFW,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,kv_cache,llama3.1-8b,1,,,,,,,,file,block,local,onprem,available,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,,,
+closed,NewFW,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,training,unet3d,6,2,1,98.45749211195111,85.0432807582894,11889.837614388982,11.61116954530174,,file,block,local,onprem,available,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,,,
+closed,NewFW,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,vector_database,milvus,0,,,,,,,,file,block,local,onprem,available,Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC,,,
+closed,OpenLake,openlake-hbv4-1c1s-final,checkpointing,llama3-8b,1,8,1,,,,,,object,object,remote_and_local,cloud,available,openlake-hbv4-1c1s-final,6.721265847525814,11.54586773861578,104.70452117919922
+closed,SAMSUNG,MEMORY_AE,checkpointing,llama3-8b,1,8,1,,,,,,file,file,local,onprem,available,MEMORY_AE,15.91369658349501,17.768773315490236,104.70452117919922
+closed,SAMSUNG,MEMORY_AE,kv_cache,llama3.1-8b,1,,,,,,,,file,file,local,onprem,available,MEMORY_AE,,,
+closed,SAMSUNG,MEMORY_AE,training,retinanet,6,40,1,87.02617251613161,17526.462240880854,5398.076692512662,5.271559270031896,,file,file,local,onprem,available,MEMORY_AE,,,
+closed,SAMSUNG,MEMORY_AE,training,unet3d,6,5,1,90.2125332923292,194.77936341169067,27231.95743331344,26.593708430970157,,file,file,local,onprem,available,MEMORY_AE,,,
+closed,SAMSUNG,MEMORY_AE,vector_database,milvus,0,,,,,,,,file,file,local,onprem,available,MEMORY_AE,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-1Tier0,checkpointing,llama3-70b,1,64,1,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-1Tier0,108.7796022777087,96.34937994892863,911.2121276855469
+closed,Suzhou_Zishan_Longlin,GalaxSphere-1Tier0,checkpointing,llama3-8b,2,8,1,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-1Tier0,61.104552410833456,95.07595164095959,104.70452117919922
+closed,Suzhou_Zishan_Longlin,GalaxSphere-1Tier0,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-1Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-1Tier0,training,retinanet,6,58,1,85.21752129397012,24909.620024881548,7672.058252692861,7.492244387395372,,file,file,remote_and_local,onprem,available,GalaxSphere-1Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-1Tier0,training,unet3d,6,26,1,93.17511137255794,1046.0972580021612,146254.0769311856,142.82624700311095,,file,file,remote_and_local,onprem,available,GalaxSphere-1Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-1Tier0,vector_database,milvus,0,,,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-1Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-2Tier0,checkpointing,llama3-405b,1,512,2,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-2Tier0,193.1774002782767,156.3314317056638,5292.248718261719
+closed,Suzhou_Zishan_Longlin,GalaxSphere-2Tier0,checkpointing,llama3-70b,1,64,2,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-2Tier0,205.01570249903548,192.8697776186547,911.2121276855469
+closed,Suzhou_Zishan_Longlin,GalaxSphere-2Tier0,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-2Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-2Tier0,training,retinanet,6,98,2,85.75751399821232,42357.78756467526,13046.020506405666,12.740254400786784,,file,file,remote_and_local,onprem,available,GalaxSphere-2Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-2Tier0,training,unet3d,6,48,2,90.85166532213339,1883.1379115023014,263280.10600742896,257.10947852287984,,file,file,remote_and_local,onprem,available,GalaxSphere-2Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-3Tier0,checkpointing,llama3-405b,1,512,3,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-3Tier0,278.5662059387463,248.75978827010493,5292.248718261719
+closed,Suzhou_Zishan_Longlin,GalaxSphere-3Tier0,checkpointing,llama3-70b,1,64,3,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-3Tier0,268.5066876911133,274.3383104925385,911.2121276855469
+closed,Suzhou_Zishan_Longlin,GalaxSphere-3Tier0,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-3Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-3Tier0,training,retinanet,6,153,3,85.0289398886973,65564.18131600162,20193.492226859984,19.720207252792953,,file,file,remote_and_local,onprem,available,GalaxSphere-3Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-3Tier0,training,unet3d,6,60,3,91.32707642920835,2366.3804728877603,330841.8878672434,323.08778112035486,,file,file,remote_and_local,onprem,available,GalaxSphere-3Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-4Tier0,checkpointing,llama3-405b,1,512,4,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-4Tier0,313.67484810321906,336.54286928623367,5292.248718261719
+closed,Suzhou_Zishan_Longlin,GalaxSphere-4Tier0,checkpointing,llama3-70b,1,64,4,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-4Tier0,288.5818278180524,353.33637539350633,911.2121276855469
+closed,Suzhou_Zishan_Longlin,GalaxSphere-4Tier0,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote_and_local,onprem,available,GalaxSphere-4Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-4Tier0,training,retinanet,6,200,4,85.26155602120818,85940.01589997794,26469.16362286489,25.848792600453994,,file,file,remote_and_local,onprem,available,GalaxSphere-4Tier0,,,
+closed,Suzhou_Zishan_Longlin,GalaxSphere-4Tier0,training,unet3d,6,80,4,91.87512315145747,3174.0172192117825,443756.9786255464,433.35642443901014,,file,file,remote_and_local,onprem,available,GalaxSphere-4Tier0,,,
+closed,TTA,seahorse-storage,checkpointing,llama3-70b,2,64,4,,,,,,file,file,remote,onprem,preview,seahorse-storage,50.31068329694689,114.73168124134213,911.2121276855469
+closed,TTA,seahorse-storage,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,preview,seahorse-storage,,,
+closed,TTA,seahorse-storage,training,unet3d,6,16,4,95.43505522495711,659.52362071741,92207.50520516025,90.04639180191431,,file,file,remote,onprem,preview,seahorse-storage,,,
+closed,TTA,seahorse-storage,vector_database,milvus,6,,,,,,,,file,file,remote,onprem,preview,seahorse-storage,,,
+closed,TuringData,TuringData_F9200_2_Clients,checkpointing,llama3-8b,2,8,2,,,,,,file,file,remote,onprem,available,TuringData_F9200_2_Clients,108.7342015896141,166.0462349669915,104.70452117919922
+closed,TuringData,TuringData_F9200_8_Clients,checkpointing,llama3-70b,2,64,8,,,,,,file,file,remote,onprem,available,TuringData_F9200_8_Clients,307.12258805134917,539.942728967935,911.2121276855469
+closed,TuringData,TuringData_F9200_8_Clients,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,available,TuringData_F9200_8_Clients,,,
+closed,TuringData,TuringData_F9200_8_Clients,training,unet3d,6,96,8,95.65503623214474,3966.0739204981182,554493.8349146331,541.4978856588214,,file,file,remote,onprem,available,TuringData_F9200_8_Clients,,,
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,checkpointing,llama3-1t,1,1024,16,,,,,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,293.60833390938285,461.98101865907176,15426.589264154434
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,checkpointing,llama3-405b,1,512,16,,,,,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,293.7616977464435,459.73118689563336,5292.248718261719
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,checkpointing,llama3-70b,1,64,16,,,,,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,268.97113920212735,459.1889908932503,911.2121276855469
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,,,
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,training,retinanet,6,688,16,85.59663337742138,296770.46791734215,91404.05655592066,89.26177398039127,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,,,
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,training,unet3d,6,80,16,96.25608493006206,3326.0067787339044,465006.52551140543,454.10793506973187,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem,,,
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem_MI355,training,retinanet,6,736,16,85.9931695256223,319071.6102236296,98272.71463679578,95.96944788749587,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem_MI355,,,
+closed,UBIX,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_1hosts256GBMem,checkpointing,llama3-8b,1,8,1,,,,,,file,file,remote,onprem,available,UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_1hosts256GBMem,39.43581878409022,45.815375362359205,104.70452117919922
+closed,XSKY,XSKY_AIMesh_3StorageNode_1Client,training,unet3d,6,12,1,95.73055874055697,496.0811234462088,69356.73164001439,67.73118324220155,,file,file,remote,onprem,available,XSKY_AIMesh_3StorageNode_1Client,,,
+closed,XSKY,XSKY_AIMesh_3StorageNode_7Client,training,unet3d,6,70,7,92.09811530350575,2782.6638088841582,389042.15039757686,379.92397499763365,,file,file,remote,onprem,available,XSKY_AIMesh_3StorageNode_7Client,,,
+closed,XSKY,XSKY_AIMesh_3StorageNode_8Client,checkpointing,llama3-405b,1,512,8,,,,,,file,file,remote,onprem,available,XSKY_AIMesh_3StorageNode_8Client,106.71045063970064,379.9247243354464,5292.248718261719
+closed,XSKY,XSKY_AIMesh_3StorageNode_8Client,checkpointing,llama3-70b,2,64,8,,,,,,file,file,remote,onprem,available,XSKY_AIMesh_3StorageNode_8Client,63.369146754331645,219.96080675055077,911.2121276855469
+closed,XSKY,XSKY_AIMesh_3StorageNode_8Client,checkpointing,llama3-8b,2,8,8,,,,,,file,file,remote,onprem,available,XSKY_AIMesh_3StorageNode_8Client,13.079841317317483,34.668375810323475,104.70452117919922
+closed,XSKY,XSKY_AIMesh_3StorageNode_9Client,training,retinanet,6,675,9,93.99661533752156,319605.1976912453,98437.05733372831,96.12993880246906,,file,file,remote,onprem,available,XSKY_AIMesh_3StorageNode_9Client,,,
+closed,YanRongTech,YanRongTech_F9000X_2_Clients,checkpointing,llama3-8b,2,8,2,,,,,,file,file,remote,onprem,available,YanRongTech_F9000X_2_Clients,105.60610099955227,168.9382329688966,104.70452117919922
+closed,YanRongTech,YanRongTech_F9000X_8_Clients,checkpointing,llama3-70b,2,64,8,,,,,,file,file,remote,onprem,available,YanRongTech_F9000X_8_Clients,313.7451931640422,538.7362280013265,911.2121276855469
+closed,YanRongTech,YanRongTech_F9000X_9_Clients,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,onprem,available,YanRongTech_F9000X_9_Clients,,,
+closed,YanRongTech,YanRongTech_F9000X_9_Clients,training,unet3d,6,99,9,93.16093877056913,3983.410164097518,556917.6021940986,543.8648458926745,,file,file,remote,onprem,available,YanRongTech_F9000X_9_Clients,,,
+closed,ZettaLane,mayanas-lustre-gcp_c3hc176_c3d90_1c_4s,training,unet3d,6,3,1,92.1377521210903,119.34540453471664,16685.59193964339,16.294523378557997,,file,file,remote,cloud,available,mayanas-lustre-gcp_c3hc176_c3d90_1c_4s,,,
+closed,ZettaLane,mayanas-lustre-gcp_c4hc144_c3d90_1c_2s,checkpointing,llama3-8b,2,8,1,,,,,,file,file,remote,cloud,available,mayanas-lustre-gcp_c4hc144_c3d90_1c_2s,14.430101911790317,10.397719908266286,104.70452117919922
+closed,ZettaLane,mayanas-lustre-gcp_c4hc144_c3d90_1c_2s,kv_cache,llama3.1-8b,1,,,,,,,,file,file,remote,cloud,available,mayanas-lustre-gcp_c4hc144_c3d90_1c_2s,,,
+closed,ZettaLane,mayanas-lustre-gcp_c4hc144_c3d90_2c_4s,checkpointing,llama3-70b,2,64,2,,,,,,file,file,remote,cloud,available,mayanas-lustre-gcp_c4hc144_c3d90_2c_4s,32.42064381943113,21.200583766601035,911.2121276855469
+closed,ZettaLane,mayascale-gcp_c4hc144_n2hc64_1c_2s,kv_cache,llama3.1-8b,0,,,,,,,,file,block,remote,cloud,available,mayascale-gcp_c4hc144_n2hc64_1c_2s,,,
+closed,ZettaLane,mayascale-gcp_c4hc144_n2hc64_1c_2s,training,retinanet,6,72,1,86.68421472392815,31437.627677779084,9682.65716737032,9.455719890010078,,file,block,remote,cloud,available,mayascale-gcp_c4hc144_n2hc64_1c_2s,,,
+closed,ZettaLane,mayascale-gcp_c4hc192_n2hc64_1c_4s,training,unet3d,6,4,1,93.25250012926668,161.0639888004911,22518.236070954288,21.990464913041297,,file,block,remote,cloud,available,mayascale-gcp_c4hc192_n2hc64_1c_4s,,,
+open,FarmGPU,potato8-qlc-np6-open-tnf-70b,kv_cache,llama3.1-70b-instruct,1,,,,,,,,file,file,local,onprem,available,potato8-qlc-np6-open-tnf-70b,,,
+open,FarmGPU,potato8-qlc-np6-open-tnf-u400,kv_cache,llama3.1-8b,1,,,,,,,,file,file,local,onprem,available,potato8-qlc-np6-open-tnf-u400,,,

json/accelerators_summary.json ADDED Viewed

The diff for this file is too large to render. See raw diff

json/claims.json ADDED Viewed

The diff for this file is too large to render. See raw diff

json/mac_run.jsonl ADDED Viewed

	@@ -0,0 +1,17 @@

+{"kind": "env", "at": "2026-10-28T11:30:58Z", "loadavg": "{ 3.72 4.13 4.30 }", "powermode": "2", "thermal": "Note: No thermal warning level has been recorded Note: No performance warning level has been recorded Note: No CPU power status has been recorded "}
+{"kind": "cpu_triad", "threads": 6, "gib_per_array": 1.0, "reps": 15, "best_gbs": 361.9, "median_gbs": 359.5, "worst_gbs": 158.1}
+{"kind": "cpu_triad", "threads": 6, "gib_per_array": 1.0, "reps": 15, "best_gbs": 362.9, "median_gbs": 361.1, "worst_gbs": 158.9}
+{"kind": "cpu_triad", "threads": 6, "gib_per_array": 1.0, "reps": 15, "best_gbs": 361.8, "median_gbs": 360.9, "worst_gbs": 168.2}
+{"kind": "cpu_triad", "threads": 12, "gib_per_array": 1.0, "reps": 15, "best_gbs": 302.1, "median_gbs": 280.6, "worst_gbs": 187.8}
+{"kind": "cpu_triad", "threads": 12, "gib_per_array": 1.0, "reps": 15, "best_gbs": 345.4, "median_gbs": 286.5, "worst_gbs": 188.9}
+{"kind": "cpu_triad", "threads": 12, "gib_per_array": 1.0, "reps": 15, "best_gbs": 287.0, "median_gbs": 278.7, "worst_gbs": 191.2}
+{"kind": "cpu_triad", "threads": 18, "gib_per_array": 1.0, "reps": 15, "best_gbs": 361.0, "median_gbs": 349.1, "worst_gbs": 177.3}
+{"kind": "cpu_triad", "threads": 18, "gib_per_array": 1.0, "reps": 15, "best_gbs": 358.2, "median_gbs": 350.4, "worst_gbs": 206.1}
+{"kind": "cpu_triad", "threads": 18, "gib_per_array": 1.0, "reps": 15, "best_gbs": 359.3, "median_gbs": 350.1, "worst_gbs": 202.5}
+{"kind": "gpu_copyk", "mib_per_buffer": 1024, "reps": 20, "best_gbs": 557.7, "median_gbs": 556.4, "worst_gbs": 421.4}
+{"kind": "gpu_triad", "mib_per_buffer": 1024, "reps": 20, "best_gbs": 560.1, "median_gbs": 559.4, "worst_gbs": 538.0}
+{"kind": "gpu_copyk", "mib_per_buffer": 1024, "reps": 20, "best_gbs": 557.9, "median_gbs": 556.9, "worst_gbs": 533.3}
+{"kind": "gpu_triad", "mib_per_buffer": 1024, "reps": 20, "best_gbs": 560.2, "median_gbs": 559.5, "worst_gbs": 547.6}
+{"kind": "gpu_copyk", "mib_per_buffer": 1024, "reps": 20, "best_gbs": 557.8, "median_gbs": 555.9, "worst_gbs": 537.6}
+{"kind": "gpu_triad", "mib_per_buffer": 1024, "reps": 20, "best_gbs": 560.2, "median_gbs": 558.9, "worst_gbs": 545.0}
+{"kind": "env_after", "loadavg": "{ 3.50 4.08 4.28 }", "at": "2026-10-28T11:31:02Z"}

json/storage_v30_results.json ADDED Viewed

	@@ -0,0 +1,2809 @@

+{
+ "source": "https://github.com/mlcommons/storage_results_v3.0",
+ "licence": "Apache-2.0",
+ "rows": [
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "aws_c5n18xlarge_i3en24xlarge_8c_10s_8a",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 8,
+   "hosts": 8,
+   "au_pct": 92.63041795978016,
+   "samples_per_s": 319.9516722797155,
+   "io_MiB_per_s": 44732.20451913498,
+   "io_GiB_per_s": 43.683793475717756,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aws_c5n18xlarge_i3en24xlarge_8c_10s_8a"
+  },
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "aws_c7i48xlarge_i3en12xlarge_1c_1s_32a",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 32,
+   "hosts": 1,
+   "au_pct": 90.37935652358514,
+   "samples_per_s": 14564.284495896372,
+   "io_MiB_per_s": 4485.738399449544,
+   "io_GiB_per_s": 4.380603905712445,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aws_c7i48xlarge_i3en12xlarge_1c_1s_32a"
+  },
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "aws_c7i48xlarge_i3en12xlarge_1c_1s_8a",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 8,
+   "hosts": 1,
+   "au_pct": 93.13657320344252,
+   "samples_per_s": 3748.975740110628,
+   "io_MiB_per_s": 1154.6687680234033,
+   "io_GiB_per_s": 1.1276062187728548,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aws_c7i48xlarge_i3en12xlarge_1c_1s_8a"
+  },
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "aws_m7i48xlarge_i3en12xlarge_32c_32s_512p",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 2,
+   "accelerators": 512,
+   "hosts": 32,
+   "write_GiB_per_s": 147.02534700759554,
+   "read_GiB_per_s": 124.34398888798535,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aws_m7i48xlarge_i3en12xlarge_32c_32s_512p"
+  },
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "aws_r5n24xlarge_i3en12xlarge_1c_1s_8h",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 4.643973844563218,
+   "read_GiB_per_s": 5.67713748443572,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aws_r5n24xlarge_i3en12xlarge_1c_1s_8h"
+  },
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "aws_r5n24xlarge_i3en12xlarge_4c_10s_64p",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 4,
+   "write_GiB_per_s": 34.494582339203355,
+   "read_GiB_per_s": 34.0776448290619,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aws_r5n24xlarge_i3en12xlarge_4c_10s_64p"
+  },
+  {
+   "division": "closed",
+   "submitter": "Alluxio",
+   "system": "rigA_c5n18xl",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 1,
+   "hosts": 1,
+   "au_pct": 97.5410258394956,
+   "samples_per_s": 42.1183576653727,
+   "io_MiB_per_s": 5888.536151954891,
+   "io_GiB_per_s": 5.750523585893448,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "rigA_c5n18xl"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-1280TiB-16-Standard_E192is_v6",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 16,
+   "write_GiB_per_s": 105.50381161813687,
+   "read_GiB_per_s": 123.59892086203688,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-1280TiB-16-Standard_E192is_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-160TiB-2-Standard_E192is_v6",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 2,
+   "write_GiB_per_s": 13.34064865889334,
+   "read_GiB_per_s": 14.818591941034665,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-160TiB-2-Standard_E192is_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-160TiB-3-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 3,
+   "hosts": 3,
+   "au_pct": 98.20510827238871,
+   "samples_per_s": 127.21469802722649,
+   "io_MiB_per_s": 17785.79198991944,
+   "io_GiB_per_s": 17.3689374901557,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-160TiB-3-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-2400TiB-46-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 46,
+   "hosts": 46,
+   "au_pct": 92.77719557885989,
+   "samples_per_s": 1842.725581446979,
+   "io_MiB_per_s": 257630.08830241417,
+   "io_GiB_per_s": 251.59188310782633,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-2400TiB-46-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 2,
+   "accelerators": 1024,
+   "hosts": 128,
+   "write_GiB_per_s": 642.2033253302709,
+   "read_GiB_per_s": 489.5958768747438,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 2,
+   "accelerators": 512,
+   "hosts": 128,
+   "write_GiB_per_s": 456.97121734565565,
+   "read_GiB_per_s": 438.1095590967043,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-4096TiB-128-Standard_E104is_v5"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-4096TiB-70-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 70,
+   "hosts": 70,
+   "au_pct": 91.8768355574808,
+   "samples_per_s": 2776.9432224078223,
+   "io_MiB_per_s": 388242.35947163624,
+   "io_GiB_per_s": 379.14292917151977,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-4096TiB-70-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-125-800TiB-16-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 16,
+   "hosts": 16,
+   "au_pct": 95.99397939512524,
+   "samples_per_s": 663.1905469956098,
+   "io_MiB_per_s": 92720.17543146123,
+   "io_GiB_per_s": 90.54704631978636,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-125-800TiB-16-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-20-25632TiB-70-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 70,
+   "hosts": 70,
+   "au_pct": 91.60451752198989,
+   "samples_per_s": 2768.7111586189276,
+   "io_MiB_per_s": 387091.44077696075,
+   "io_GiB_per_s": 378.01898513375073,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-20-25632TiB-70-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-20-4800TiB-16-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 16,
+   "hosts": 16,
+   "au_pct": 95.78464205233865,
+   "samples_per_s": 661.743771559797,
+   "io_MiB_per_s": 92517.90283751945,
+   "io_GiB_per_s": 90.34951448976508,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-20-4800TiB-16-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-20-960TiB-3-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 3,
+   "hosts": 3,
+   "au_pct": 98.17320004551571,
+   "samples_per_s": 127.17207287441481,
+   "io_MiB_per_s": 17779.832599116304,
+   "io_GiB_per_s": 17.363117772574515,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-20-960TiB-3-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-250-400TiB-16-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 16,
+   "hosts": 16,
+   "au_pct": 96.0973913226102,
+   "samples_per_s": 663.9028653332451,
+   "io_MiB_per_s": 92819.76412663762,
+   "io_GiB_per_s": 90.64430090491955,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-250-400TiB-16-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-250-80TiB-3-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 3,
+   "hosts": 3,
+   "au_pct": 98.225381172166,
+   "samples_per_s": 127.24153035136264,
+   "io_MiB_per_s": 17789.543397131754,
+   "io_GiB_per_s": 17.37260097376148,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-250-80TiB-3-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-40-2400TiB-16-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 16,
+   "hosts": 16,
+   "au_pct": 94.37566077762118,
+   "samples_per_s": 652.0094255355602,
+   "io_MiB_per_s": 91156.95118468511,
+   "io_GiB_per_s": 89.02046014129405,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-40-2400TiB-16-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-40-480TiB-3-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 3,
+   "hosts": 3,
+   "au_pct": 97.87151696783833,
+   "samples_per_s": 126.78031014244225,
+   "io_MiB_per_s": 17725.060543934633,
+   "io_GiB_per_s": 17.309629437436165,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-40-480TiB-3-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-500-200TiB-16-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 16,
+   "hosts": 16,
+   "au_pct": 95.68081570244276,
+   "samples_per_s": 661.0190848423504,
+   "io_MiB_per_s": 92416.58492839227,
+   "io_GiB_per_s": 90.25057121913308,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-500-200TiB-16-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Azure",
+   "system": "AMLFS-Durable-Premium-500-40TiB-3-Standard_D192ds_v6",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 3,
+   "hosts": 3,
+   "au_pct": 98.19383172702176,
+   "samples_per_s": 127.19967213116847,
+   "io_MiB_per_s": 17783.691230605502,
+   "io_GiB_per_s": 17.366885967388185,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "AMLFS-Durable-Premium-500-40TiB-3-Standard_D192ds_v6"
+  },
+  {
+   "division": "closed",
+   "submitter": "Everpure",
+   "system": "Everpure_FBEXA_file_32hosts_10-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 1,
+   "accelerators": 1024,
+   "hosts": 32,
+   "write_GiB_per_s": 327.64719805437477,
+   "read_GiB_per_s": 365.1207598586278,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Everpure_FBEXA_file_32hosts_10-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c"
+  },
+  {
+   "division": "closed",
+   "submitter": "Everpure",
+   "system": "Everpure_FBEXA_file_32hosts_15-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 1,
+   "accelerators": 1024,
+   "hosts": 32,
+   "write_GiB_per_s": 484.09584386379964,
+   "read_GiB_per_s": 439.87114303619046,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Everpure_FBEXA_file_32hosts_15-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c"
+  },
+  {
+   "division": "closed",
+   "submitter": "Everpure",
+   "system": "Everpure_FBEXA_file_32hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 1,
+   "accelerators": 1024,
+   "hosts": 32,
+   "write_GiB_per_s": 655.6004361087373,
+   "read_GiB_per_s": 533.756143699845,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Everpure_FBEXA_file_32hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c"
+  },
+  {
+   "division": "closed",
+   "submitter": "Everpure",
+   "system": "Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 2,
+   "accelerators": 1024,
+   "hosts": 32,
+   "write_GiB_per_s": 877.5237989244531,
+   "read_GiB_per_s": 588.2768047121776,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c"
+  },
+  {
+   "division": "closed",
+   "submitter": "Everpure",
+   "system": "Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 32,
+   "write_GiB_per_s": 744.2387157606029,
+   "read_GiB_per_s": 833.016210806981,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Everpure_FBEXA_file_32hosts_30-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c"
+  },
+  {
+   "division": "closed",
+   "submitter": "Everpure",
+   "system": "Everpure_FBEXA_file_51hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 2,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Everpure_FBEXA_file_51hosts_20-DN-DNOS-1.1.0-4845253c3899_30-S500R2-blades-MDN-4.8.3.exa-4845253c"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "hickory",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 1,
+   "accelerators": 1024,
+   "hosts": 15,
+   "write_GiB_per_s": 169.67001937204188,
+   "read_GiB_per_s": 252.1079035740201,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "hickory"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "hickory",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 15,
+   "write_GiB_per_s": 165.2149160928629,
+   "read_GiB_per_s": 283.65453242888884,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "hickory"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "hickory",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "hickory"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "hickory",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 75,
+   "hosts": 15,
+   "au_pct": 91.96409502125466,
+   "samples_per_s": 2978.8917029423214,
+   "io_MiB_per_s": 416476.6258195245,
+   "io_GiB_per_s": 406.7154549018794,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "hickory"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "potato1-qlc-closed-tcm",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "potato1-qlc-closed-tcm"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "potato8-qlc-closed-tcm",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "potato8-qlc-closed-tcm"
+  },
+  {
+   "division": "closed",
+   "submitter": "FarmGPU",
+   "system": "potato8-qlc-np2-closed-tcm",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "potato8-qlc-np2-closed-tcm"
+  },
+  {
+   "division": "closed",
+   "submitter": "HPE",
+   "system": "K3000-kvcache",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "K3000-kvcache"
+  },
+  {
+   "division": "closed",
+   "submitter": "HPE",
+   "system": "e2000",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 16,
+   "write_GiB_per_s": 48.18024515327689,
+   "read_GiB_per_s": 110.22794945362989,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "e2000"
+  },
+  {
+   "division": "closed",
+   "submitter": "HPE",
+   "system": "e2000-unet3d",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 30,
+   "hosts": 15,
+   "au_pct": 94.55619792991766,
+   "samples_per_s": 1225.0496535146963,
+   "io_MiB_per_s": 171273.27779430087,
+   "io_GiB_per_s": 167.25906034599694,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "e2000-unet3d"
+  },
+  {
+   "division": "closed",
+   "submitter": "HPE",
+   "system": "k3000",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 70.26223153852276,
+   "read_GiB_per_s": 83.74542808097782,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "k3000"
+  },
+  {
+   "division": "closed",
+   "submitter": "HPE",
+   "system": "k3000-unet3d",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 60,
+   "hosts": 15,
+   "au_pct": 97.73403055373916,
+   "samples_per_s": 2532.6903360986644,
+   "io_MiB_per_s": 354093.54572448286,
+   "io_GiB_per_s": 345.7944782465653,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "k3000-unet3d"
+  },
+  {
+   "division": "closed",
+   "submitter": "HolmesAI",
+   "system": "StorMeshAI",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 20,
+   "hosts": 1,
+   "au_pct": 85.61695963222215,
+   "samples_per_s": 8623.848804413834,
+   "io_MiB_per_s": 2656.1091788550175,
+   "io_GiB_per_s": 2.593856619975603,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "StorMeshAI"
+  },
+  {
+   "division": "closed",
+   "submitter": "HolmesAI",
+   "system": "StorMeshAI",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 4,
+   "hosts": 1,
+   "au_pct": 90.8331384551441,
+   "samples_per_s": 156.88409272415188,
+   "io_MiB_per_s": 21933.84792000856,
+   "io_GiB_per_s": 21.41977335938336,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "StorMeshAI"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_10clients",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_10clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_10clients",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 50,
+   "hosts": 10,
+   "au_pct": 91.20689736028054,
+   "samples_per_s": 1969.219883543951,
+   "io_MiB_per_s": 275315.1622749616,
+   "io_GiB_per_s": 268.8624631591422,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_10clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_10clients_checkpointing",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 10,
+   "write_GiB_per_s": 195.86855929267355,
+   "read_GiB_per_s": 269.9591183910151,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_10clients_checkpointing"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_1client",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 33.031401803633365,
+   "read_GiB_per_s": 45.011544959951635,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_1client"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_KH-50000_10clients",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 10,
+   "write_GiB_per_s": 202.40849606043844,
+   "read_GiB_per_s": 368.79434917650207,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_KH-50000_10clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_KH-50000_10clients",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_KH-50000_10clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "InspurData",
+   "system": "InspurData_InDataOS_AS13000G7_KH-50000_10clients_Unet3D",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 70,
+   "hosts": 10,
+   "au_pct": 90.6251395681733,
+   "samples_per_s": 2739.234101116296,
+   "io_MiB_per_s": 382970.27536646323,
+   "io_GiB_per_s": 373.99440953756175,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "InspurData_InDataOS_AS13000G7_KH-50000_10clients_Unet3D"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-12n-oci-bm-denseio-e5-128-16client",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 2,
+   "accelerators": 1024,
+   "hosts": 16,
+   "write_GiB_per_s": 133.26620428586173,
+   "read_GiB_per_s": 136.54350674628003,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-12n-oci-bm-denseio-e5-128-16client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-12n-oci-bm-denseio-e5-128-16client",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 2,
+   "accelerators": 512,
+   "hosts": 16,
+   "write_GiB_per_s": 114.53121018453218,
+   "read_GiB_per_s": 123.2881347348884,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-12n-oci-bm-denseio-e5-128-16client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-12n-oci-bm-denseio-e5-128-1client",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 3.196488777385128,
+   "read_GiB_per_s": 8.334717693512255,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-12n-oci-bm-denseio-e5-128-1client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-12n-oci-bm-denseio-e5-128-20client-unet3d",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 20,
+   "hosts": 20,
+   "au_pct": 98.01536666413432,
+   "samples_per_s": 846.5426709017167,
+   "io_MiB_per_s": 118354.49903773211,
+   "io_GiB_per_s": 115.58056546653526,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-12n-oci-bm-denseio-e5-128-20client-unet3d"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-12n-oci-bm-denseio-e5-128-8client",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 27.724210916654336,
+   "read_GiB_per_s": 78.12166947859546,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-12n-oci-bm-denseio-e5-128-8client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-aws-i8ge-150g-4client-m8gn24xl-unet3d",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 8,
+   "hosts": 4,
+   "au_pct": 98.37396216523095,
+   "samples_per_s": 339.92408507480917,
+   "io_MiB_per_s": 47524.5326464581,
+   "io_GiB_per_s": 46.41067641255674,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-aws-i8ge-150g-4client-m8gn24xl-unet3d"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-aws-i8ge-150g-5client-m8gn24xl",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 255,
+   "hosts": 5,
+   "au_pct": 88.76981905247644,
+   "samples_per_s": 114053.24730387353,
+   "io_MiB_per_s": 35127.920712964136,
+   "io_GiB_per_s": 34.30461007125404,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-aws-i8ge-150g-5client-m8gn24xl"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-gcp-z3-200g-8client-n2d96",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 8,
+   "hosts": 8,
+   "au_pct": 97.88103113179451,
+   "samples_per_s": 338.0440734228716,
+   "io_MiB_per_s": 47261.68962046727,
+   "io_GiB_per_s": 46.153993769987565,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-gcp-z3-200g-8client-n2d96"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-oci-bm-denseio-e5-128-1client",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 3.0202352379882162,
+   "read_GiB_per_s": 6.992930457598599,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-oci-bm-denseio-e5-128-1client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-oci-bm-denseio-e5-128-5client-retinanet",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 190,
+   "hosts": 5,
+   "au_pct": 94.06365488499786,
+   "samples_per_s": 90053.51900612167,
+   "io_MiB_per_s": 27736.10528722767,
+   "io_GiB_per_s": 27.086040319558272,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-oci-bm-denseio-e5-128-5client-retinanet"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-oci-bm-denseio-e5-128-5client-unet3d",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 5,
+   "hosts": 5,
+   "au_pct": 98.86264187253634,
+   "samples_per_s": 213.4684552990488,
+   "io_MiB_per_s": 29844.865422277908,
+   "io_GiB_per_s": 29.14537638894327,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-oci-bm-denseio-e5-128-5client-unet3d"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-oci-bm-denseio-e5-128-8client",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 22.14796788267677,
+   "read_GiB_per_s": 33.28990554938302,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-oci-bm-denseio-e5-128-8client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 2,
+   "accelerators": 1024,
+   "hosts": 8,
+   "write_GiB_per_s": 33.61921399099397,
+   "read_GiB_per_s": 34.19594857148875,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 2,
+   "accelerators": 512,
+   "hosts": 8,
+   "write_GiB_per_s": 33.30672578549643,
+   "read_GiB_per_s": 33.076730530779834,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-3n-oci-bm-denseio-e5-128-8client-405b-1t"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-6n-oci-bm-denseio-e5-128-10client-retinanet",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 370,
+   "hosts": 10,
+   "au_pct": 89.25373008545304,
+   "samples_per_s": 166384.46279087468,
+   "io_MiB_per_s": 51245.715093185914,
+   "io_GiB_per_s": 50.04464364568937,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-6n-oci-bm-denseio-e5-128-10client-retinanet"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-6n-oci-bm-denseio-e5-128-10client-unet3d",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 10,
+   "hosts": 10,
+   "au_pct": 98.74707647368561,
+   "samples_per_s": 426.4364823599876,
+   "io_MiB_per_s": 59619.76634605895,
+   "io_GiB_per_s": 58.22242807232319,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-6n-oci-bm-denseio-e5-128-10client-unet3d"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-6n-oci-bm-denseio-e5-128-1client",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 3.3973937004064005,
+   "read_GiB_per_s": 11.08184250902501,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-6n-oci-bm-denseio-e5-128-1client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-6n-oci-bm-denseio-e5-128-8client",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 26.541383802027422,
+   "read_GiB_per_s": 56.79712068487546,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-6n-oci-bm-denseio-e5-128-8client"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 2,
+   "accelerators": 1024,
+   "hosts": 8,
+   "write_GiB_per_s": 64.50496710106077,
+   "read_GiB_per_s": 70.00709074663709,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t"
+  },
+  {
+   "division": "closed",
+   "submitter": "NVIDIA",
+   "system": "aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 2,
+   "accelerators": 512,
+   "hosts": 8,
+   "write_GiB_per_s": 62.027056682203764,
+   "read_GiB_per_s": 63.80655689735703,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "aistore-6n-oci-bm-denseio-e5-128-8client-405b-1t"
+  },
+  {
+   "division": "closed",
+   "submitter": "Nebius",
+   "system": "nebius-s3-ckpt-llama3-70b",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 19.45538992050427,
+   "read_GiB_per_s": 38.274839274839515,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "s3",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "nebius-s3-ckpt-llama3-70b"
+  },
+  {
+   "division": "closed",
+   "submitter": "Nebius",
+   "system": "nebius-s3-ckpt-llama3-8b",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 1,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 2.8118349276291923,
+   "read_GiB_per_s": 7.143715604587631,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "s3",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "nebius-s3-ckpt-llama3-8b"
+  },
+  {
+   "division": "closed",
+   "submitter": "Nebius",
+   "system": "nebius-s3-retinanet-32nx24-cm64",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 768,
+   "hosts": 32,
+   "au_pct": 88.42844382305479,
+   "samples_per_s": 341881.3336218671,
+   "io_MiB_per_s": 105298.0135560201,
+   "io_GiB_per_s": 102.83009136330088,
+   "uri_scheme": "s3",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "nebius-s3-retinanet-32nx24-cm64"
+  },
+  {
+   "division": "closed",
+   "submitter": "Nebius",
+   "system": "nebius-s3-unet3d-21n",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 21,
+   "hosts": 21,
+   "au_pct": 92.47475811918585,
+   "samples_per_s": 838.2511080278598,
+   "io_MiB_per_s": 117195.26182039271,
+   "io_GiB_per_s": 114.44849787147726,
+   "uri_scheme": "s3",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "nebius-s3-unet3d-21n"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 3.076140817854296,
+   "read_GiB_per_s": 4.983509448925309,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 1,
+   "hosts": 1,
+   "au_pct": 97.74196560717776,
+   "samples_per_s": 42.2128683477068,
+   "io_MiB_per_s": 5901.749619918002,
+   "io_GiB_per_s": 5.763427363201174,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D5-P5336_61TB_Gen4_NVMe_QLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 8.63733219350259,
+   "read_GiB_per_s": 12.999883786501638,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 2,
+   "hosts": 1,
+   "au_pct": 98.45749211195111,
+   "samples_per_s": 85.0432807582894,
+   "io_MiB_per_s": 11889.837614388982,
+   "io_GiB_per_s": 11.61116954530174,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "NewFW",
+   "system": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC",
+   "category": "vector_database",
+   "workload": "milvus",
+   "runs": 0,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "Solidigm_D7-PS1010_7TB_Gen5_NVMe_TLC"
+  },
+  {
+   "division": "closed",
+   "submitter": "OpenLake",
+   "system": "openlake-hbv4-1c1s-final",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 1,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 6.721265847525814,
+   "read_GiB_per_s": 11.54586773861578,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "object",
+   "product_API": "object",
+   "storage_location": "remote_and_local",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "openlake-hbv4-1c1s-final"
+  },
+  {
+   "division": "closed",
+   "submitter": "SAMSUNG",
+   "system": "MEMORY_AE",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 1,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 15.91369658349501,
+   "read_GiB_per_s": 17.768773315490236,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "MEMORY_AE"
+  },
+  {
+   "division": "closed",
+   "submitter": "SAMSUNG",
+   "system": "MEMORY_AE",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "MEMORY_AE"
+  },
+  {
+   "division": "closed",
+   "submitter": "SAMSUNG",
+   "system": "MEMORY_AE",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 40,
+   "hosts": 1,
+   "au_pct": 87.02617251613161,
+   "samples_per_s": 17526.462240880854,
+   "io_MiB_per_s": 5398.076692512662,
+   "io_GiB_per_s": 5.271559270031896,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "MEMORY_AE"
+  },
+  {
+   "division": "closed",
+   "submitter": "SAMSUNG",
+   "system": "MEMORY_AE",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 5,
+   "hosts": 1,
+   "au_pct": 90.2125332923292,
+   "samples_per_s": 194.77936341169067,
+   "io_MiB_per_s": 27231.95743331344,
+   "io_GiB_per_s": 26.593708430970157,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "MEMORY_AE"
+  },
+  {
+   "division": "closed",
+   "submitter": "SAMSUNG",
+   "system": "MEMORY_AE",
+   "category": "vector_database",
+   "workload": "milvus",
+   "runs": 0,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "MEMORY_AE"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-1Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 1,
+   "write_GiB_per_s": 108.7796022777087,
+   "read_GiB_per_s": 96.34937994892863,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-1Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-1Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 61.104552410833456,
+   "read_GiB_per_s": 95.07595164095959,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-1Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-1Tier0",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-1Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-1Tier0",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 58,
+   "hosts": 1,
+   "au_pct": 85.21752129397012,
+   "samples_per_s": 24909.620024881548,
+   "io_MiB_per_s": 7672.058252692861,
+   "io_GiB_per_s": 7.492244387395372,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-1Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-1Tier0",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 26,
+   "hosts": 1,
+   "au_pct": 93.17511137255794,
+   "samples_per_s": 1046.0972580021612,
+   "io_MiB_per_s": 146254.0769311856,
+   "io_GiB_per_s": 142.82624700311095,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-1Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-1Tier0",
+   "category": "vector_database",
+   "workload": "milvus",
+   "runs": 0,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-1Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-2Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 2,
+   "write_GiB_per_s": 193.1774002782767,
+   "read_GiB_per_s": 156.3314317056638,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-2Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-2Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 2,
+   "write_GiB_per_s": 205.01570249903548,
+   "read_GiB_per_s": 192.8697776186547,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-2Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-2Tier0",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-2Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-2Tier0",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 98,
+   "hosts": 2,
+   "au_pct": 85.75751399821232,
+   "samples_per_s": 42357.78756467526,
+   "io_MiB_per_s": 13046.020506405666,
+   "io_GiB_per_s": 12.740254400786784,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-2Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-2Tier0",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 48,
+   "hosts": 2,
+   "au_pct": 90.85166532213339,
+   "samples_per_s": 1883.1379115023014,
+   "io_MiB_per_s": 263280.10600742896,
+   "io_GiB_per_s": 257.10947852287984,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-2Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-3Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 3,
+   "write_GiB_per_s": 278.5662059387463,
+   "read_GiB_per_s": 248.75978827010493,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-3Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-3Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 3,
+   "write_GiB_per_s": 268.5066876911133,
+   "read_GiB_per_s": 274.3383104925385,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-3Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-3Tier0",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-3Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-3Tier0",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 153,
+   "hosts": 3,
+   "au_pct": 85.0289398886973,
+   "samples_per_s": 65564.18131600162,
+   "io_MiB_per_s": 20193.492226859984,
+   "io_GiB_per_s": 19.720207252792953,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-3Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-3Tier0",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 60,
+   "hosts": 3,
+   "au_pct": 91.32707642920835,
+   "samples_per_s": 2366.3804728877603,
+   "io_MiB_per_s": 330841.8878672434,
+   "io_GiB_per_s": 323.08778112035486,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-3Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-4Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 4,
+   "write_GiB_per_s": 313.67484810321906,
+   "read_GiB_per_s": 336.54286928623367,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-4Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-4Tier0",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 4,
+   "write_GiB_per_s": 288.5818278180524,
+   "read_GiB_per_s": 353.33637539350633,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-4Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-4Tier0",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-4Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-4Tier0",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 200,
+   "hosts": 4,
+   "au_pct": 85.26155602120818,
+   "samples_per_s": 85940.01589997794,
+   "io_MiB_per_s": 26469.16362286489,
+   "io_GiB_per_s": 25.848792600453994,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-4Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "Suzhou_Zishan_Longlin",
+   "system": "GalaxSphere-4Tier0",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 80,
+   "hosts": 4,
+   "au_pct": 91.87512315145747,
+   "samples_per_s": 3174.0172192117825,
+   "io_MiB_per_s": 443756.9786255464,
+   "io_GiB_per_s": 433.35642443901014,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote_and_local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "GalaxSphere-4Tier0"
+  },
+  {
+   "division": "closed",
+   "submitter": "TTA",
+   "system": "seahorse-storage",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 4,
+   "write_GiB_per_s": 50.31068329694689,
+   "read_GiB_per_s": 114.73168124134213,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "seahorse-storage"
+  },
+  {
+   "division": "closed",
+   "submitter": "TTA",
+   "system": "seahorse-storage",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "seahorse-storage"
+  },
+  {
+   "division": "closed",
+   "submitter": "TTA",
+   "system": "seahorse-storage",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 16,
+   "hosts": 4,
+   "au_pct": 95.43505522495711,
+   "samples_per_s": 659.52362071741,
+   "io_MiB_per_s": 92207.50520516025,
+   "io_GiB_per_s": 90.04639180191431,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "seahorse-storage"
+  },
+  {
+   "division": "closed",
+   "submitter": "TTA",
+   "system": "seahorse-storage",
+   "category": "vector_database",
+   "workload": "milvus",
+   "runs": 6,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "preview",
+   "submission_name": "seahorse-storage"
+  },
+  {
+   "division": "closed",
+   "submitter": "TuringData",
+   "system": "TuringData_F9200_2_Clients",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 2,
+   "write_GiB_per_s": 108.7342015896141,
+   "read_GiB_per_s": 166.0462349669915,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "TuringData_F9200_2_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "TuringData",
+   "system": "TuringData_F9200_8_Clients",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 307.12258805134917,
+   "read_GiB_per_s": 539.942728967935,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "TuringData_F9200_8_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "TuringData",
+   "system": "TuringData_F9200_8_Clients",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "TuringData_F9200_8_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "TuringData",
+   "system": "TuringData_F9200_8_Clients",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 96,
+   "hosts": 8,
+   "au_pct": 95.65503623214474,
+   "samples_per_s": 3966.0739204981182,
+   "io_MiB_per_s": 554493.8349146331,
+   "io_GiB_per_s": 541.4978856588214,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "TuringData_F9200_8_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem",
+   "category": "checkpointing",
+   "workload": "llama3-1t",
+   "runs": 1,
+   "accelerators": 1024,
+   "hosts": 16,
+   "write_GiB_per_s": 293.60833390938285,
+   "read_GiB_per_s": 461.98101865907176,
+   "checkpoint_GB": 15426.589264154434,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 16,
+   "write_GiB_per_s": 293.7616977464435,
+   "read_GiB_per_s": 459.73118689563336,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 1,
+   "accelerators": 64,
+   "hosts": 16,
+   "write_GiB_per_s": 268.97113920212735,
+   "read_GiB_per_s": 459.1889908932503,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 688,
+   "hosts": 16,
+   "au_pct": 85.59663337742138,
+   "samples_per_s": 296770.46791734215,
+   "io_MiB_per_s": 91404.05655592066,
+   "io_GiB_per_s": 89.26177398039127,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 80,
+   "hosts": 16,
+   "au_pct": 96.25608493006206,
+   "samples_per_s": 3326.0067787339044,
+   "io_MiB_per_s": 465006.52551140543,
+   "io_GiB_per_s": 454.10793506973187,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem_MI355",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 736,
+   "hosts": 16,
+   "au_pct": 85.9931695256223,
+   "samples_per_s": 319071.6102236296,
+   "io_MiB_per_s": 98272.71463679578,
+   "io_GiB_per_s": 95.96944788749587,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_16hosts256GBMem_MI355"
+  },
+  {
+   "division": "closed",
+   "submitter": "UBIX",
+   "system": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_1hosts256GBMem",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 1,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 39.43581878409022,
+   "read_GiB_per_s": 45.815375362359205,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "UbiPower18000_16x15.36TB_NVMe_4xNDR400_3xStorageNode_1hosts256GBMem"
+  },
+  {
+   "division": "closed",
+   "submitter": "XSKY",
+   "system": "XSKY_AIMesh_3StorageNode_1Client",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 12,
+   "hosts": 1,
+   "au_pct": 95.73055874055697,
+   "samples_per_s": 496.0811234462088,
+   "io_MiB_per_s": 69356.73164001439,
+   "io_GiB_per_s": 67.73118324220155,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "XSKY_AIMesh_3StorageNode_1Client"
+  },
+  {
+   "division": "closed",
+   "submitter": "XSKY",
+   "system": "XSKY_AIMesh_3StorageNode_7Client",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 70,
+   "hosts": 7,
+   "au_pct": 92.09811530350575,
+   "samples_per_s": 2782.6638088841582,
+   "io_MiB_per_s": 389042.15039757686,
+   "io_GiB_per_s": 379.92397499763365,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "XSKY_AIMesh_3StorageNode_7Client"
+  },
+  {
+   "division": "closed",
+   "submitter": "XSKY",
+   "system": "XSKY_AIMesh_3StorageNode_8Client",
+   "category": "checkpointing",
+   "workload": "llama3-405b",
+   "runs": 1,
+   "accelerators": 512,
+   "hosts": 8,
+   "write_GiB_per_s": 106.71045063970064,
+   "read_GiB_per_s": 379.9247243354464,
+   "checkpoint_GB": 5292.248718261719,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "XSKY_AIMesh_3StorageNode_8Client"
+  },
+  {
+   "division": "closed",
+   "submitter": "XSKY",
+   "system": "XSKY_AIMesh_3StorageNode_8Client",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 63.369146754331645,
+   "read_GiB_per_s": 219.96080675055077,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "XSKY_AIMesh_3StorageNode_8Client"
+  },
+  {
+   "division": "closed",
+   "submitter": "XSKY",
+   "system": "XSKY_AIMesh_3StorageNode_8Client",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 8,
+   "write_GiB_per_s": 13.079841317317483,
+   "read_GiB_per_s": 34.668375810323475,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "XSKY_AIMesh_3StorageNode_8Client"
+  },
+  {
+   "division": "closed",
+   "submitter": "XSKY",
+   "system": "XSKY_AIMesh_3StorageNode_9Client",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 675,
+   "hosts": 9,
+   "au_pct": 93.99661533752156,
+   "samples_per_s": 319605.1976912453,
+   "io_MiB_per_s": 98437.05733372831,
+   "io_GiB_per_s": 96.12993880246906,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "XSKY_AIMesh_3StorageNode_9Client"
+  },
+  {
+   "division": "closed",
+   "submitter": "YanRongTech",
+   "system": "YanRongTech_F9000X_2_Clients",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 2,
+   "write_GiB_per_s": 105.60610099955227,
+   "read_GiB_per_s": 168.9382329688966,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "YanRongTech_F9000X_2_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "YanRongTech",
+   "system": "YanRongTech_F9000X_8_Clients",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 8,
+   "write_GiB_per_s": 313.7451931640422,
+   "read_GiB_per_s": 538.7362280013265,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "YanRongTech_F9000X_8_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "YanRongTech",
+   "system": "YanRongTech_F9000X_9_Clients",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "YanRongTech_F9000X_9_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "YanRongTech",
+   "system": "YanRongTech_F9000X_9_Clients",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 99,
+   "hosts": 9,
+   "au_pct": 93.16093877056913,
+   "samples_per_s": 3983.410164097518,
+   "io_MiB_per_s": 556917.6021940986,
+   "io_GiB_per_s": 543.8648458926745,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "YanRongTech_F9000X_9_Clients"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayanas-lustre-gcp_c3hc176_c3d90_1c_4s",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 3,
+   "hosts": 1,
+   "au_pct": 92.1377521210903,
+   "samples_per_s": 119.34540453471664,
+   "io_MiB_per_s": 16685.59193964339,
+   "io_GiB_per_s": 16.294523378557997,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayanas-lustre-gcp_c3hc176_c3d90_1c_4s"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayanas-lustre-gcp_c4hc144_c3d90_1c_2s",
+   "category": "checkpointing",
+   "workload": "llama3-8b",
+   "runs": 2,
+   "accelerators": 8,
+   "hosts": 1,
+   "write_GiB_per_s": 14.430101911790317,
+   "read_GiB_per_s": 10.397719908266286,
+   "checkpoint_GB": 104.70452117919922,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayanas-lustre-gcp_c4hc144_c3d90_1c_2s"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayanas-lustre-gcp_c4hc144_c3d90_1c_2s",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayanas-lustre-gcp_c4hc144_c3d90_1c_2s"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayanas-lustre-gcp_c4hc144_c3d90_2c_4s",
+   "category": "checkpointing",
+   "workload": "llama3-70b",
+   "runs": 2,
+   "accelerators": 64,
+   "hosts": 2,
+   "write_GiB_per_s": 32.42064381943113,
+   "read_GiB_per_s": 21.200583766601035,
+   "checkpoint_GB": 911.2121276855469,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayanas-lustre-gcp_c4hc144_c3d90_2c_4s"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayascale-gcp_c4hc144_n2hc64_1c_2s",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 0,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayascale-gcp_c4hc144_n2hc64_1c_2s"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayascale-gcp_c4hc144_n2hc64_1c_2s",
+   "category": "training",
+   "workload": "retinanet",
+   "runs": 6,
+   "accelerators": 72,
+   "hosts": 1,
+   "au_pct": 86.68421472392815,
+   "samples_per_s": 31437.627677779084,
+   "io_MiB_per_s": 9682.65716737032,
+   "io_GiB_per_s": 9.455719890010078,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayascale-gcp_c4hc144_n2hc64_1c_2s"
+  },
+  {
+   "division": "closed",
+   "submitter": "ZettaLane",
+   "system": "mayascale-gcp_c4hc192_n2hc64_1c_4s",
+   "category": "training",
+   "workload": "unet3d",
+   "runs": 6,
+   "accelerators": 4,
+   "hosts": 1,
+   "au_pct": 93.25250012926668,
+   "samples_per_s": 161.0639888004911,
+   "io_MiB_per_s": 22518.236070954288,
+   "io_GiB_per_s": 21.990464913041297,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "block",
+   "storage_location": "remote",
+   "deployment": "cloud",
+   "availability": "available",
+   "submission_name": "mayascale-gcp_c4hc192_n2hc64_1c_4s"
+  },
+  {
+   "division": "open",
+   "submitter": "FarmGPU",
+   "system": "potato8-qlc-np6-open-tnf-70b",
+   "category": "kv_cache",
+   "workload": "llama3.1-70b-instruct",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "potato8-qlc-np6-open-tnf-70b"
+  },
+  {
+   "division": "open",
+   "submitter": "FarmGPU",
+   "system": "potato8-qlc-np6-open-tnf-u400",
+   "category": "kv_cache",
+   "workload": "llama3.1-8b",
+   "runs": 1,
+   "uri_scheme": "",
+   "benchmark_API": "file",
+   "product_API": "file",
+   "storage_location": "local",
+   "deployment": "onprem",
+   "availability": "available",
+   "submission_name": "potato8-qlc-np6-open-tnf-u400"
+  }
+ ]
+}