AI BriefingHugging Face Releases ROCm Kernel Build Tool
Key point
Hugging Face has released a tool that makes it easy to build and share ROCm kernels for AMD GPUs.
Details
Hugging Face's kernels library and kernel-builder make it easy to build custom kernels supporting various accelerator backends such as CUDA, ROCm, Metal, and XPU, and integrate them with PyTorch.
It focuses on improving kernel portability and reproducibility by resolving the complex CMake/Nix configurations, compiler errors, and ABI issues that arise when building custom kernels.
The guide uses the RadeonFlow GEMM kernel, optimized for the AMD Instinct MI300X GPU, as an example. This kernel uses the FP8 low-precision format to deliver high throughput and low memory bandwidth, and is a high-performance matrix multiplication implementation that won the Grand Prize at the AMD Developer Challenge 2025.
Using this tool, developers can structure and build the kernels they write, and easily share them with developers around the world through kernels-community.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
OpenArch Released: A Collection of PyTorch Implementations for Modern LLM Architectures
ZeroModels Released, Compatible with JAX, PyTorch, and TensorFlowGoogle Releases Gemini-Based AI Agent Tools to Accelerate TPU Development