←back to thread

AMD funded a drop-in CUDA implementation built on ROCm: It's now open-source

(www.phoronix.com)
1045 points mfiguiere | 1 comments | 12 Feb 24 14:00 UTC | HN request time: 0.22s | source
Show context
sam_goody ◴[12 Feb 24 14:34 UTC] No.39345188[source]▶
I don't really follow this, but isn't it a bad sign for ROCm that, for example, ZLUDA + Blender 4's CUDA back-end delivers better performance than the native Radeon HIP back-end?
replies(4): >>39345214 #>>39345404 #>>39345464 #>>39345623 #
1. KeplerBoy ◴[12 Feb 24 14:59 UTC] No.39345464[source]▶
Surely this can be attributed to Blender's HIP code just being suboptimal because nobody really cares about it. By extension nobody cares about it because performance is suboptimal.

It's AMDs job to break that circle.