llama.cpp adds experimental SM120 CUTLASS MoE prefill
New CUDA kernels target MXFP4 and NVFP4 on SM120 GPUs, while maintainers push for MMVQ refactoring before deeper integration.
By tensor#moe
llama.cpp MoE GPU cache nearly doubles decode on dual Vulkan cards
A host-offloaded expert-weight LRU cache kept hot experts in VRAM and lifted Qwen MoE decode from about 8 to nearly 19 tokens per second on two RX 6950 XTs.
By tensor