[DOC] Update sycl_ext_oneapi_num_compute_units.asciidoc (https://github.com/intel/llvm/issues/21408)#21429
Conversation
Add a table entry for AMD GPU and a reference to streaming multiprocessors
There was a problem hiding this comment.
I think it is better to link to primary sources. I don't know what's the right source, but here are some suggestions:
Intel: https://www.intel.com/content/www/us/en/docs/oneapi/optimization-guide-gpu/2025-2/intel-xe-gpu-architecture.html#INTEL-R-XE-GPU-BUILDING-BLOCKS
NVIDIA: https://docs.nvidia.com/deeplearning/performance/dl-performance-gpu-background/index.html#gpu-arch
AMD: https://rocm.docs.amd.com/projects/HIP/en/latest/understand/hardware_implementation.html#hierarchical-organization
Update the table with the suggestions
sycl/doc/extensions/supported/sycl_ext_oneapi_num_compute_units.asciidoc
Outdated
Show resolved
Hide resolved
|
@YuriPlyakhin do you have any more concerns with this PR? |
|
@intel/llvm-gatekeepers please consider merging |
2 similar comments
|
@intel/llvm-gatekeepers please consider merging |
|
@intel/llvm-gatekeepers please consider merging |
|
@intel/llvm-gatekeepers I think this is ready to merge. |
This commit will be authored by an anonymous email, which goes against LLVM contribution policies. See https://llvm.org/docs/DeveloperPolicy.html#email-addresses
|
It is disabled. |
|
Thanks, I can see your email now |
catching up. no concerns. |
Add a table entry for AMD GPU and a reference to streaming multiprocessors