onnxruntime
[CPU/MLAS] Add QMoE shared layout INT4/INT3/FP4 AVX-512 & AVX2 kernels, FileExpertStore, and direct CPU QMoE
#33450
Merged
Overview
Commits
7
Changes
View On GitHub

[CPU/MLAS] Add QMoE shared layout INT4/INT3/FP4 AVX-512 & AVX2 kernels, FileExpertStore, and direct CPU QMoE #33450

tianleiwu merged 7 commits into main from tlwu/20261019/qmoe-offload-p1-cpu-shared-layout
tianleiwu
github-advanced-security
github-advanced-security commented on 2026-10-19
tianleiwu tianleiwu force pushed from a0d08bfe to 7a920381 22 days ago
mirounga mirounga requested a review from mirounga mirounga 22 days ago
tianleiwu tianleiwu force pushed from 7a920381 to 7e3016c3 21 days ago
tianleiwu tianleiwu force pushed from 7e3016c3 to 5f1dc948 21 days ago
tianleiwu tianleiwu force pushed from 5f1dc948 to c4535f38 20 days ago
tianleiwu tianleiwu force pushed from c4535f38 to 3c007ec3 20 days ago
tianleiwu tianleiwu marked this pull request as ready for review 19 days ago
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 19 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-10-21
tianleiwu tianleiwu force pushed from 3c007ec3 to 0cc4f641 19 days ago
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 18 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-10-23
github-advanced-security
github-advanced-security commented on 2026-10-23
tianleiwu tianleiwu requested a review from apsonawane apsonawane 17 days ago
tianleiwu tianleiwu requested a review from kunal-vaishnavi kunal-vaishnavi 17 days ago
tianleiwu tianleiwu requested a review from titaiwangms titaiwangms 17 days ago
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 17 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-10-23
tianleiwu tianleiwu force pushed from dd0da0f7 to ea94e4be 17 days ago
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 16 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-10-24
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 15 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-10-26
tianleiwu
tianleiwu tianleiwu force pushed from 631a7e92 to 129ace23 13 days ago
BBaditi
BBaditi commented on 2026-10-27
BBaditi
BBaditi commented on 2026-10-28
BBaditi
BBaditi commented on 2026-10-28
kunal-vaishnavi
kunal-vaishnavi commented on 2026-10-28
tianleiwu tianleiwu added qmoe_offload
tianleiwu tianleiwu force pushed from d0b033d9 to a3bc0acc 11 days ago
tianleiwu tianleiwu changed the title [CPU/MLAS] Add QMoE shared 4-bit expert layouts, AVX-512 kernels, and FileExpertStore [CPU/MLAS] Add QMoE shared layout INT4 AVX-512/AVX2 kernels, FileExpertStore, and direct CPU QMoE 11 days ago
github-actions
github-actions commented on 2026-10-29
github-advanced-security
github-advanced-security commented on 2026-10-29
tianleiwu tianleiwu changed the title [CPU/MLAS] Add QMoE shared layout INT4 AVX-512/AVX2 kernels, FileExpertStore, and direct CPU QMoE [CPU/MLAS] Add QMoE shared layout INT4/INT3/FP4 AVX-512 & AVX2 kernels, FileExpertStore, and direct CPU QMoE 11 days ago
github-actions
github-actions commented on 2026-10-29
github-advanced-security
github-advanced-security commented on 2026-10-29
tianleiwu tianleiwu force pushed from 3e56894f to 8cbf133c 11 days ago
tianleiwu tianleiwu force pushed from 8cbf133c to 3046801e 11 days ago
tianleiwu tianleiwu force pushed from 8a383788 to 49fb21b3 11 days ago
tianleiwu tianleiwu force pushed from 49fb21b3 to 8a383788 11 days ago
tianleiwu [CPU/MLAS] Add QMoE shared layout INT4 AVX-512/AVX2 kernels, FileExpe…
05a22ab8
tianleiwu [CPU/MLAS] Add QMoE shared layout INT3 AVX-512/AVX2 kernels and direc…
cf003059
tianleiwu [CPU/MLAS] Add QMoE shared layout NVFP4/MXFP4 AVX-512 kernels and sha…
d86249ab
tianleiwu fix(mlas): restore QMoE FP16 rounding helpers
e45bd154
tianleiwu fix(test,docs): skip unavailable QUICK_BUILD CUDA BF16 QMoE runner an…
5e14bce6
tianleiwu tianleiwu force pushed from 39594fe1 to 5e14bce6 11 days ago
baijumeswani
baijumeswani commented on 2026-10-30
tianleiwu Fix QMoE shared-layout numerics and add deferred scaling option
860e5e8f
tianleiwu fix(mlas): avoid MSVC C4702 in QMoE INT4 round helpers
8bdf4852
baijumeswani
baijumeswani approved these changes on 2026-10-30
tianleiwu tianleiwu enabled auto-merge (squash) 10 days ago
tianleiwu tianleiwu merged c574f2d4 into main 10 days ago
tianleiwu tianleiwu deleted the tlwu/20261019/qmoe-offload-p1-cpu-shared-layout branch 10 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
qmoe_offload
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum