MTP tuning can unlock 50%–100% more throughput on some local model setups
bradrlaw · reddit · 2026-07-25
The author benchmarks MTP tuning across several model families on a local setup with a P100 and two V100s, and finds that leaving nmax at defaults can waste a lot of performance.
- Different model families respond very differently to the same settings.
- On some models and GPU/card combinations, proper tuning can recover 50–100% more throughput than the default configuration.
- The post includes benchmark scripts, VRAM impact notes, and a GitHub write-up with results for models such as Gemma and Qwen.
More from Infra
- Anthropic S-1: training infra edge cut per-task cost 40% and doubled revenue growth — BenBajarin · 2026-10-26
- Rewriting scaling law study: optimal synthetic data share rises from 10% to 32% with compute — heghbalz · 2026-10-26
- US states halt new data centers as 72% of Americans oppose local builds — YvesMulkers · 2026-10-26
- $5M Seed Round Can't Afford One DeepSeek Training Run: GPU-less AI Startup's Three Paths — FinanceYF5 · 2026-10-26
- HBM demand shifts: low-stack inference plans shelved, Rubin to go 12-high only — BenBajarin · 2026-10-26
- CircuitHub unveils The Grid, a factory-scale robot making PCBs in the US twice as fast as Shenzhen — garrytan · 2026-10-26