logoalt Hacker News

Matrix Core Programming on AMD CDNA Architecture

37 points • by salykova • last Tuesday at 3:57 PM • 5 comments • view on HN

Comments

phkahler • today at 5:30 AM

So from CDNA3 to 4 they doubled fp16 and fp8 performance but cut fp32 and fp64 by half?

Wonder why the regression on non-AI workloads?

➕ show 3 replies