whn09 ad23d2d870 Neuron MoE capacity path: all-reduce the overflow flag; bench prompts differ per rank (#26)
The overflow decision was rank-local, and the dense rerun has collectives of
its own: one rank rerunning left the others waiting in its all-gather (NRT
'Collective operation pending'). The flag is now max-all-reduced over the EP
group. The bench gave every rank the same prompt, which routes 64 copies of
each token to the same experts (overflow on every layer); real requests
differ, so the prompts now take the rank in their seed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-23 17:06:47 +08:00
..
DeepEP@93eb6eb238 Neuron MoE capacity path: all-reduce the overflow flag; bench prompts differ per rank (#26) 2026-10-23 17:06:47 +08:00
DeepEP-Ascend@3b25377d04 initial commit 2026-10-20 12:09:42 +08:00
DeepGEMM@1fc9c66975 initial commit 2026-10-20 12:09:42 +08:00
DeepGEMM-Ascend@3a35326271 initial commit 2026-10-20 12:09:42 +08:00
DeepSelect@bfa4507d93 initial commit 2026-10-20 12:09:42 +08:00
FlashMLA@2e5429fc56 initial commit 2026-10-20 12:09:42 +08:00