blogvLLM Blog·23 Oct 20265.1
vLLM serves DeepSeek-V4.1-Flash on one GB300 NVL72 rack at 954K tokens per second per GPU using prefill and decode disaggregation and wide expert parallelism, 7.9 times its single-node result.
paperHugging Face Daily Papers·20 Oct 20263.9
Meta Expert shares experts across layers to cut MoE storage and retrieval.
paperHugging Face Daily Papers·22 Oct 20263.7
IQuest-Q1 is a 320B sparse MoE model built for agentic CLI tasks.
blogvLLM Blog·21 Oct 20263.2
vLLM runs Recursive Flint, a 40B MoE with 2.5B active parameters, on one GPU at about 500 tokens per second with 100K context.