Great work to @vllm_project team and @nvidia on smooth, out-of-the-box day 0 @MiniMax_AI M3 experience with @inferact EAGLE3 spec decode. Here are the details of ongoing M3 workstream:
NVIDIA, Inferact and SemiAnalysis are working hard on enabling disaggregated inferencing (PR
π£: MiniMax M3 has landed, joining models like DeepSeek V4 and Kimi-K2.6 at the frontier of open agentic models β and NVIDIA Blackwell is already delivering leading performance on it.
NVIDIA Blackwell Ultra delivers up to 5x higher AI factory throughput than NVIDIA Hopper on
