NHacker Next
  • new
  • past
  • show
  • ask
  • show
  • jobs
  • submit
▲Writing high-performance matrix multiplication kernels for Blackwell (docs.jax.dev)
33 points by lairv 4 days ago | 1 comment
arjvik 1 hours ago [-]
The interesting part is this is done in Pallas!

Seems like the Pallas of old has completely been upgraded

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact
Rendered at 21:59:46 GMT+0000 (Coordinated Universal Time) with Vercel.