Measured on B200: nvfp4 is ~9% more efficient than mxfp4 with higher accuracy — pick nvfp4 on Blackwell

StasBekman · x · 2026-10-03

Stas Bekman measured nvfp4 to be about 9% more efficient than mxfp4 on B200 using MAMF metrics, with higher accuracy as well.

His takeaway: on Blackwell and newer GPUs, choose nvfp4 over mxfp4. He also links to an explanation of fp4 formats.

Original post →

More from Infra

Infra channel →