Skip to content

[GPU] Integrate grouped gemm to support moe_3gemm #34974

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
riverlijunjie wants to merge 4 commits into
base: master
Choose a base branch
from
Draft

[GPU] Integrate grouped gemm to support moe_3gemm #34974

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
35d2fdb
Enable grouped_gemm for moe_3gemm
riverlijunjie Mar 27, 2026
9bd5207
[GPU] moe_3gemm support grouped_gemm for prefill phase
riverlijunjie Mar 9, 2026
dcef1c4
[Test] Enable max_variable_dim parameter for grouped_gemm
riverlijunjie Mar 31, 2026
e83f73f
Merge branch 'master' into river/grouped_gemm_integration
peterchen-intel Apr 1, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions src/plugins/intel_gpu/src/graph/impls/ocl_v2/moe/moe_3gemm_base.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,7 @@ namespace ov::intel_gpu::ocl {
#define MOE_INTERNAL_BUFFER_TOKEN_LEN_PER_ACTIVATED_EXPERT 11 // token len (input gather tokens) for each activated expert
#define MOE_INTERNAL_BUFFER_TOKEN_IDX_PER_EXPERT 12 // token idx per expert
#define MOE_INTERNAL_BUFFER_ACTUAL_USED_EXPERT_NUM 13 // num_actual_used_experts
#define MOE_INTERNAL_BUFFER_GROUPED_OFFSETS 14 // int32_t cumulative end-offsets per expert for OneDNN grouped GEMM

#define ENABLE_MOE_MICRO_GEMM_POST_PROC_SILU_MUL 1

Expand Down
Loading
Loading