At FP8 precision, AI training typically achieves only around 35% model floating-point operations utilization (MFU), meaning GPUs are only used at peak potential about a third of the time, with the rest spent waiting for data movement between caches or other GPUs.

factualpending

Speaker

Unidentified Speaker — The Engineering Unlocks Behind DeepSeek | YC Decoded [4Tmn-XP93m4]

Evidence Quote

“at fp8 it is typical to only see around 35% model flops utilization or mfu meaning gpus are only being utilized at Peak potential about a third of the time”

Source

The Engineering Unlocks Behind DeepSeek | YC Decoded— Y Combinator
Created: 8/11/2026, 6:17:44 AM

Your notes

Loading notes...