Coding Tools

llama.cpp b11270 updates ROCm and CUDA builds

llama.cpp ReleasesSeptember 30, 2026high confidence

Why it matters

→ROCm 10.0 support allows AMD GPU users to run newer drivers without custom compilation.
→CUDA 13.4 updates ensure compatibility with the latest NVIDIA driver releases.
→HIP optimizations may provide minor performance improvements on specific AMD architectures.

llama.cpp has released build b11270, updating its precompiled binaries to support ROCm 10.0 and CUDA 13.4 across Linux and Windows platforms. The release includes specific HIP optimizations for packed byte subtraction operations to improve efficiency on AMD GPUs. Additionally, builds for openEuler with ACL Graph support have been disabled in this version. This update ensures compatibility with the latest driver stacks for both NVIDIA and AMD hardware.

Read original

llama.cpp b11270 updates ROCm and CUDA builds

Why it matters

More from llama.cpp Releases

llama.cpp fixes MoE dispatch inefficiency on Vulkan

llama.cpp Vulkan perf boost on Intel Arc

More in Coding Tools

Claude Code v2.1.284 adds Sonnet 5.5

Claude Code v2.1.285 patches security and stability

llama.cpp b11267 adds ROCm 10 and Snapdragon support

GitHub adds repo-level Dependabot runner config