Open Source
llama.cpp b9877 Release Expands Platform Support
llama.cpp Releaseshigh confidence
Why it matters
- →Expands support for AMD GPUs with ROCm 7.2, offering alternatives to CUDA.
- →Reinforces llama.cpp's versatility across multiple operating systems and hardware.
- →Continues to serve as a flexible inference runtime for diverse environments.
The b9877 release of llama.cpp has been announced, focusing on expanding platform support rather than introducing new features. Key highlights include the addition of ROCm 7.2 support for Ubuntu x64, enhancing options for AMD GPU users. The release also maintains broad compatibility across macOS, Linux, Windows, and openEuler platforms. While some features like KleidiAI for macOS are disabled, the update reinforces llama.cpp's role as a flexible tool for AI inference across various systems.
Read originalMore from llama.cpp Releases
Coding Toolscoding
llama.cpp b10993 adds ROCm 10 and CUDA 13 builds
This release quietly extends llama.cpp's hardware support to the latest driver stacks, adding official binaries for ROCm 10.0 and CUDA 13 across Linux and Windows. For AMD users, this means native compatibility with newer GPU architectures without manual compilation tweaks, while NVIDIA users gain access to the latest CUDA runtime optimizations. The inclusion of WebGPU in CI signals ongoing work toward browser-based inference, though it remains a background effort for now. There are no new model formats or quantization methods here, just broader infrastructure coverage that keeps llama.cpp relevant as hardware evolves.
llama.cpp Releases
Coding Toolscoding
llama.cpp fixes Metal NaN bug for large activations
A critical precision bug in llama.cpp’s Apple Silicon backend has been patched, resolving total inference failures on models with high-activation ranges like Mistral Small 4. The issue stemmed from f16 saturation during matrix multiplication, which turned entire output tensors into NaN values for inputs exceeding ~32 tokens. By implementing an exact, power-of-two rescaling mechanism in the Metal kernel, the fix restores correctness without significant performance penalties. This ensures local inference on M-series chips remains viable for complex MoE architectures that previously crashed.
