Skip to content

Gemm reduce max#209

Merged
asroy merged 48 commits into
developfrom
gemm_reduce_max
May 20, 2022
Merged

Gemm reduce max#209
asroy merged 48 commits into
developfrom
gemm_reduce_max

Conversation

@rocking5566
Copy link
Copy Markdown
Collaborator

@rocking5566 rocking5566 commented Apr 27, 2022 •
edited
Loading

I use hip library version of atomicMax instead of buffer addressing.

rosenrodt
rosenrodt reviewed Apr 28, 2022
View reviewed changes
Comment thread example/19_gemm_softmax/gemm_softmax_xdl_fp16.cpp Outdated
Comment thread include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp Outdated
Comment thread include/ck/utility/generic_memory_space_atomic_max.hpp Outdated
Comment thread example/19_gemm_softmax/gemm_softmax_xdl_fp16.cpp Outdated
Comment thread include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp Outdated
rosenrodt
rosenrodt reviewed Apr 29, 2022
View reviewed changes
Comment thread include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp Outdated
Comment thread include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp Outdated
asroy
asroy reviewed Apr 30, 2022
View reviewed changes
Comment thread include/ck/utility/generic_memory_space_atomic_add.hpp
asroy
asroy reviewed Apr 30, 2022
View reviewed changes
Comment thread include/ck/utility/generic_memory_space_atomic_max.hpp Outdated
asroy
asroy reviewed Apr 30, 2022
View reviewed changes
Comment thread include/ck/utility/generic_memory_space_atomic_max.hpp Outdated
@asroy asroy self-requested a review April 30, 2022 04:12
asroy
asroy suggested changes Apr 30, 2022
View reviewed changes
Copy link
Copy Markdown
Contributor

@asroy asroy left a comment •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please do:

  • rebase this PR on develop, where GEMM+reduce has been updated (#213)

  • MI200 has atomic_max for fp64, please add buffer atomic max for fp64 in next PR

@asroy asroy mentioned this pull request Apr 30, 2022
Closed
rosenrodt
rosenrodt reviewed May 3, 2022
View reviewed changes
Comment thread include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp Outdated
@rocking5566 rocking5566 requested a review from rosenrodt May 3, 2022 09:04
@zjing14 zjing14 added the WIP label May 19, 2022
@rocking5566
[What] Fix naming
d11c7ec 
[Why] Prepare to add out elementop
@rocking5566 rocking5566 force-pushed the gemm_reduce_max branch 2 times, most recently from 3bde7f7 to ec22933 Compare May 19, 2022 16:05
@zjing14 zjing14 removed the WIP label May 20, 2022
asroy
asroy approved these changes May 20, 2022
View reviewed changes
Copy link
Copy Markdown
Contributor

@asroy asroy left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Thanks for the effort

@asroy asroy merged commit 0ffe956 into develop May 20, 2022
@rocking5566 rocking5566 deleted the gemm_reduce_max branch September 6, 2022 08:21
illsilin added a commit that referenced this pull request Dec 2, 2024
@illsilin
Merge pull request #209 from ROCm/andriy/merge_from_public
7d50244 
Update develop branch from public repository
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@rosenrodt rosenrodt Awaiting requested review from rosenrodt

1 more reviewer

@asroy asroy asroy approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@rocking5566 @asroy @rosenrodt @zjing14