Fast transformer inference with Metal Performance Shaders | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		Fast transformer inference with Metal Performance Shaders (explosion.ai)
		3 points by reichardt on Nov 24, 2022 \| hide \| past \| favorite \| 1 comment

reichardt on Nov 24, 2022 [–]

Seems like the Apple M1 Max is almost half as fast as a RTX 3090. Pretty cool!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact