Third Party Hardware > GPU Compute / Accelerators
AMD OpenCL / ROCm
tle:
AFAIK it's unfortunate that we won't be able to take advantage of ROCm HIP for Blender simply because ROCm's codebase is not architecture-agnostic. There has been efforts in the past to get ROCm code compile on PPC64 with little attention from AMD. Recent codebase move so quickly most patches for PPC64 would be out of sync.
I hope that Vulkan Compute would become more mature in the future. Wishful thinking I know...
Navigation
[0] Message Index
[*] Previous page
Go to full version