onnxruntime
[webgpu] Use 64 as the workgroup size of DP4AMatMulQuantize
#24129
Merged

[webgpu] Use 64 as the workgroup size of DP4AMatMulQuantize #24129

guschmue merged 12 commits into main from matmul_quantize
qjia7
qjia7 [webgpu] Use 64 as the workgroup size of DP4AMatMulQuantize
dea3b477
qjia7 qjia7 requested a review from sushraja-msft sushraja-msft 1 year ago
qjia7 qjia7 requested a review from guschmue guschmue 1 year ago
guschmue guschmue added ep:WebGPU
qjia7 simplify
54dced44
qjia7 add subgroup 32 support
20032c71
qjia7 add subgroup 16 support
3e0451df
qjia7 Merge branch 'main' into matmul_quantize
3f798032
qjia7 Merge branch 'main' into matmul_quantize
72c9ff4e
qjia7 Merge branch 'main' into matmul_quantize
7dc1cd04
qjia7 fix build error
bbc67eb9
qjia7 use template
0e952b92
qjia7 add missing barrier
7aacff93
sushraja-msft
sushraja-msft commented on 2025-08-15
sushraja-msft
sushraja-msft dismissed these changes on 2025-08-15
sushraja-msft
sushraja-msft commented on 2025-08-15
qjia7 address comments
c27fc0d4
qjia7 qjia7 dismissed their stale review via c27fc0d4 364 days ago
qjia7 Merge branch 'main' into matmul_quantize
dadaae0b
qjia7 qjia7 requested a review from sushraja-msft sushraja-msft 361 days ago
guschmue
guschmue approved these changes on 2025-08-19
guschmue
guschmue approved these changes on 2025-08-19
guschmue guschmue merged 1ad9f121 into main 359 days ago
guschmue guschmue deleted the matmul_quantize branch 359 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone