NanoGPT speedrun record slashed to 39.9 seconds
MIT's Deven Pietrzak cut the NanoGPT speedrun record from 73.9 to 39.9 seconds on 8×H100 GPUs, a 46% improvement that exceeds all previous 45 records combined, via flop-level optimizations merged as PR #360.
2026-09-29 ~ 2026-09-29 · 4 related posts
- NanoGPT speedrun record falls to 39.9s, -46% via flop-level skipping tricks — yacinelearning · 2026-09-29
- NanoGPT trained in 39.9 seconds, a new world record cutting 46% of training time — cephaloform · 2026-09-29
- NanoGPT speedrun record falls to 39.9 seconds on 8×H100, biggest single leap yet — QuintinPope5 · 2026-09-29
1 near-duplicate retellings: ziv_ravid