README.md

---
license: apache-2.0
language:
  - en
tags:
  - tensorflow
  - tensorflow-gpu
  - aarch64
  - arm64
  - linux-aarch64
  - cuda
  - cuda-12
  - cudnn-9
  - gpu
  - blackwell
  - gb10
  - sm_90
  - sm_120
  - python-3.12
  - wheel
  - selfbuilt
pretty_name: TensorFlow GPU wheels for linux_aarch64 (CUDA 12.8 / cuDNN 9.8)
size_categories:
  - n<1K
---

# TensorFlow 2.20 GPU wheel for linux_aarch64 (CUDA 12.8 / cuDNN 9.8)

Self-built `tensorflow` wheel for the platforms PyPI does **not** ship a
GPU build for. Produced by
[`scripts/build_tf_gpu_aarch64.sh`](https://github.com/Infineon/lpwwd/blob/main/scripts/build_tf_gpu_aarch64.sh)
in the LPWWD pipeline repo on an NVIDIA Spark / GB10 host.

## Why this exists

PyPI ships a CPU-only `tensorflow` wheel for `linux_aarch64`. There is no
pip-installable GPU TensorFlow on this platform/Python combo, so to get
GPU acceleration without Docker the wheel has to be built from source.
A cold from-source build is 2–4 h and ~50–80 GB of bazel artifacts; this
repo lets every other aarch64 host skip that.

## Contents

| File | Size | sha256 |
|---|---:|---|
| `tensorflow-2.20.0.dev0+selfbuilt-cp312-cp312-linux_aarch64.whl` | ~495 MiB | `6c63ce87206ac1485b5858a100f098674943098da946837b77d8d6c07a7ec35b` |
| `tensorflow-2.20.0.dev0+selfbuilt-cp312-cp312-linux_aarch64.whl.sha256` | — | sidecar |

## Build configuration

| Setting | Value |
|---|---|
| TensorFlow | `v2.20.0` |
| Python | 3.12 (cp312) |
| Platform tag | `linux_aarch64` (ARM 64-bit) |
| CUDA | 12.8 (hermetic) |
| cuDNN | 9.8 (hermetic) |
| Compute capabilities | `9.0` (Hopper) + `12.0` (Blackwell / GB10 `sm_120`) |
| Device compiler | `nvcc` |
| Host compiler | `clang-17` (via `--config=nvcc_clang`) |
| Bazel | 7.4.1 |
| Build host | NVIDIA Spark (GB10, aarch64, 20 cores, 121.7 GiB unified memory) |

This wheel will run on any `linux_aarch64` host with a CUDA-12.x driver
and a GPU of compute capability 9.0 or 12.0 (e.g. H100/H200/Hopper and
Blackwell/GB10). Other compute capabilities are not embedded — if your
device has e.g. `sm_80` you need a rebuild.

## Install

```bash
pip download \
    --no-deps \
    --dest . \
    "https://huggingface.co/datasets/infineon/tensorflow-gpu/resolve/main/tensorflow-2.20.0.dev0+selfbuilt-cp312-cp312-linux_aarch64.whl"
sha256sum -c <(echo "6c63ce87206ac1485b5858a100f098674943098da946837b77d8d6c07a7ec35b  tensorflow-2.20.0.dev0+selfbuilt-cp312-cp312-linux_aarch64.whl")
pip install --upgrade "./tensorflow-2.20.0.dev0+selfbuilt-cp312-cp312-linux_aarch64.whl"
```

Or with `huggingface_hub`:

```python
from huggingface_hub import hf_hub_download
whl = hf_hub_download(
    repo_id="infineon/tensorflow-gpu",
    repo_type="dataset",
    filename="tensorflow-2.20.0.dev0+selfbuilt-cp312-cp312-linux_aarch64.whl",
)
```

Then verify the GPU is visible:

```python
import tensorflow as tf
print(tf.__version__, tf.config.list_physical_devices("GPU"))
```

## Compatibility matrix

| Host arch | CUDA driver | GPU SM | Status |
|---|---|---|---|
| `linux_aarch64` | 12.8+ | `sm_90` (Hopper) | OK |
| `linux_aarch64` | 12.8+ | `sm_120` (Blackwell / GB10) | OK |
| `linux_aarch64` | 12.8+ | other SM | rebuild required |
| `linux_x86_64` | — | — | wrong arch; use upstream PyPI |
| `macOS` / Windows | — | — | not supported |

## Provenance

Built from the upstream `tensorflow/tensorflow` repo at tag `v2.20.0`
(no patches) using
[`scripts/build_tf_gpu_aarch64.sh`](https://github.com/Infineon/lpwwd/blob/main/scripts/build_tf_gpu_aarch64.sh).
The build script pins all toolchain versions (Bazel, CUDA, cuDNN, clang)
and is the single source of truth — re-running it on a fresh aarch64
host with `TF_VERSION=v2.20.0` reproduces this wheel bit-identically
modulo timestamps.

## License

TensorFlow itself is Apache-2.0. This dataset card is also Apache-2.0.