onnxruntime
using GPU_WARP_SIZE to make kernel portable between AMD and Nvidia GPU
#5173

Merged

using GPU_WARP_SIZE to make kernel portable between AMD and Nvidia GPU #5173

weixingzhang merged 1 commit into master from wezhan/reduce

using GPU_WARP_SIZE to make kernel portable between AMD and Nvidia GPU

bfde7750

weixingzhang requested a review 5 years ago

weixingzhang assigned

wschin 5 years ago

weixingzhang added core runtime

weixingzhang added training

wschin commented on 2020-09-18

wschin approved these changes on 2020-09-18

weixingzhang merged b49f6a5e into master 5 years ago

weixingzhang deleted the wezhan/reduce branch 5 years ago

Reviewers

wschin

Assignees

wschin

Labels

training core runtime

Milestone

No milestone