TRANSMISSION // 2026-05-29
vLLM + MiniMax-M2.7 on dual GB10
Two GB10 boxes, one unified inference pool.
vllm serve MiniMax-M2.7-AWQ --tensor-parallel-size 2Two GB10 boxes, one unified inference pool.
vllm serve MiniMax-M2.7-AWQ --tensor-parallel-size 2
// COMMS