← Performance Debugging & Benchmarking / 17

hardNewNVIDIADeepSeekTogether AI

MoE decode slowed after the traffic mix changed. Are hot experts or communication causing the delay?

Compare expert counts with rank-level completion times. Explain why the longest combine span can occur on a lightly loaded rank, then test a placement change against a communication control.

Updated Oct 2026 · Worked answers on inference mechanisms, with calculations, implementation details, and follow-up questions.

Compare expert counts with rank-level completion times. Explain why the longest combine span can occur on a lightly loaded rank, then test a placement change against a communication control.

10 answers per topic instead of 5, plus saved progress and bookmarks · costs nothingor unlock all 136 remaining answers · ₹2,000 / $25

Every worked answer and derivation in all 12 topics, every concept page, and the full practice-test bank.

6 months · One payment · No auto-renewal

UP NEXT ON YOUR JOURNEY

Next in this trackPrefill/decode disaggregation increased latency. Is KV transfer, admission, or token delivery on the critical path?Next in this trackImage requests OOM while the KV pool is mostly free. Separate encoder peaks from decoder memory and choose a fix.Next in this trackA router raises prefix-cache hits but worsens tail latency. Replay its decisions and find the missing state.

FEDITOR'S NOTE

Balanced token counts do not establish balanced kernel times. Ask which per-layer measurements would distinguish expert shape effects, rank arrival skew and transport progress.

DISCUSSION · 0

Nobody has commented on this answer. Add a correction or another way to work it.