DINOv2-Base (ONNX) β Renesas X5H
Not a classifier. DINOv2 is a self-supervised embedding backbone (ViT-Base). Its output is a
last_hidden_statefeature/embedding tensor, not class logits β do not treat this model as an ImageNet classifier.
Introduction
This repository hosts DINOv2 (ViT-Base) targeting the Renesas R-Car X5H platform for image feature-extraction (embedding) inference on the NPX6 NPU.
- Model Architecture: DINOv2 β self-supervised Vision Transformer (ViT-Base backbone)
- Source Model: facebook/dinov2-base β checkpoint
dinov2_base - Task: Image Feature Extraction β produces a
last_hidden_stateembedding tensor, suitable as input to downstream retrieval, clustering, or fine-tuned classification/segmentation heads - Backbone: ViT-Base
- Parameters: 86M
Deployment Flow
The FP32 ONNX model is auto-cast to INT8 by the Renesas MWMX toolchain at compile time β no separate quantization step is required.
dinov2_base_model_int8_fromHF_fp32.onnx (FP32)
β
βββΆ MWMX Runtime βββΆ INT8 auto-cast βββΆ NPX6 NPU βββΆ last_hidden_state
Provided Artifacts
| Artifact | Status | Notes |
|---|---|---|
| FP32 (ONNX) | β Published | fp32/dinov2_base_model_int8_fromHF_fp32.onnx β auto-cast to INT8 by the MWMX toolchain at compile time (see Deployment Flow above); no separate INT8 file is shipped |
Performance
Measured on Renesas R-Car X5H via the MWMX runtime (APM50 ship-performance CI pipeline).
Benchmark configuration: Single NPU Β· Batch size: 1 Β· Input resolution: not available from source data β TBD
| Parameters | Runtime | Precision | Device | Latency (ms) | Type |
|---|---|---|---|---|---|
| 86M | MWMX Runtime | INT8 (auto) | X5H Β· 1Γ NPU Β· 1 Core Β· 850 MHz | 530.03538 | Measured |
| 86M | MWMX Runtime | INT8 (auto) | X5H Β· 1Γ NPU Β· 12 Cores Β· 850 MHz | 321.912461 | Measured |
Accuracy
TBD β not applicable in the usual classification-accuracy sense. As an embedding backbone, quality would typically be evaluated via downstream task performance (e.g. k-NN classification, retrieval) or embedding similarity to the FP32 reference β not yet measured/published for this repo.
ONNX Runtime β Renesas EP v2.1.0 (MWMX 2.2 ENG4)
Measured on R-Car X5H with ONNX Runtime + Renesas Execution Provider (INT8 auto-cast); nodes the NPU cannot run fall back to the CPU EP. Batch size 1, 1 NPU. Source: ORT Renesas EP test report, status 2026-09-30. This section supersedes earlier ORT figures in this card.
| Component | AI Cores | Latency (ms) | Throughput (fps) | NPU Inference (ms) | NPU % | CPU + Overhead % | Portable pkg NPU-only (ms) |
|---|---|---|---|---|---|---|---|
| Full model | 1 | 451.443 | 2.2 | 449.585 | 99.6 % | 0.4 % | 528.402 |
| Full model | 12 | 271.338 | 3.7 | 268.613 | 99.0 % | 1.0 % | 260.76 |
12-core latency is not always lower than 1-core: small models are dominated by CPU-side overhead.
Graph partitioning (NPU vs CPU nodes)
| Component | Total Nodes | NPU Nodes | CPU Nodes (Q/DQ inserted) | NPU % | CPU % |
|---|---|---|---|---|---|
| Full model | 320 | 318 | 2 (2) | 99.4 % | 0.6 % |
Compile notes (from the test workbook)
Subgraph config:
default_settings: input_resolution: [1, 3, 518, 518]
Runtime Details
MWMX Runtime
- Engine: Renesas MWMX (Middleware MX) native inference runtime
- Input format: FP32 ONNX (compiled by the MWMX toolchain)
- NPU execution precision: INT8 (auto-cast by MWMX toolchain)
- Execution target: NPX6-48K NPU on R-Car X5H
- Output:
last_hidden_stateβ patch/token embedding tensor, not class logits
Prerequisites
To run inference on Renesas R-Car X5H, you need:
- Renesas R-Car X5H board with NPX6 NPU
- Renesas MWMX Runtime
- Hugging Face CLI to download the model
Download
hf download Renesas/DINOv2-Base-ONNX --repo-type=model --include "fp32/*"
Benchmark Methodology
- HIL runs: Hardware-in-the-loop β measured on physical R-Car X5H silicon via the MWMX
runtime (
metawaremx_runtimeCI pipeline, "APM50" ship-performance target) - Precision: FP32 ONNX input; INT8 execution (auto-cast by MWMX)
- Slices: results reported for both 1 AI core and 12 AI cores per NPU instance
Model tree for Renesas/DINOv2-Base-ONNX
Base model
facebook/dinov2-base