GitHub1 min read
b11618: HIP: use 64-row MMQ tiles for some quant types on RDNA3.5 (#28195)
LLM inference in C/C++. Contribute to ggml-org/llama.cpp development by creating an account on GitHub.
By liangliangchangllama.cpp
Read the full story
Originally published by llama.cpp on Oct 14, 2026.
Comments
Sign in to join the discussion.