New Nvidia driver adds 16bit float support to OpenCL [pdf]

by zbendefyon 2/15/2024, 9:38:13 AM with 5 comments
  • by Keltesethon 2/15/2024, 10:16:34 AM

    Can someone explain why they update to Clang 7, that was released 19 September 2018?

  • by peppermint_gumon 2/17/2024, 9:30:29 AM

    It doesn't seem it's new. The release notes for the 511.79 driver from February 2022 say the same thing:

    https://us.download.nvidia.com/Windows/511.79/511.79-win11-w...

  • by zbendefyon 2/15/2024, 9:38:13 AM

    It seems they are finally enable the 16bit float extension in their OpenCL implementation.

  • by Aardwolfon 2/17/2024, 8:37:45 AM

    Is this the IEEE float16 type or the bfloat16 type? It doesn't say

  • by bbcc90on 2/17/2024, 6:44:19 AM

    any idea why this took them so long?