Choose the right MoE token dispatcher (`alltoall`, DeepEP, or HybridEP) for the hardware, EP degree, and optimization stage. Summarizes patterns from DSV3, Qwen3, Qwen3-Next, and VLM bring-up work.
安装
npx skills add NVIDIA/skills --skill nemo-mbridge-perf-moe-dispatcher-selectionnpx skills add NVIDIA/skills --skill nemo-mbridge-perf-moe-dispatcher-selection -a claude-code -gnpx skills add NVIDIA/skills --skill nemo-mbridge-perf-moe-dispatcher-selection -a codex -ggemini skills install https://github.com/NVIDIA/skills.git --path skills/nemo-mbridge-perf-moe-dispatcher-selectiongh skill install NVIDIA/skills nemo-mbridge-perf-moe-dispatcher-selectiongit clone --depth 1 https://github.com/NVIDIA/skills /tmp/NVIDIA__skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/NVIDIA__skills/skills/nemo-mbridge-perf-moe-dispatcher-selection ~/.claude/skills/nemo-mbridge-perf-moe-dispatcher-selection # Claude Code
cp -R /tmp/NVIDIA__skills/skills/nemo-mbridge-perf-moe-dispatcher-selection ~/.agents/skills/nemo-mbridge-perf-moe-dispatcher-selection # Codex, Cursor, Gemini CLI, Copilot, OpenCode