Skip to content
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

Jump to bottom

[IE CLDNN] Fully connected MMAD kernel optimizations #2115

Merged
merged 2 commits into from
Sep 10, 2020
Merged

[IE CLDNN] Fully connected MMAD kernel optimizations #2115

merged 2 commits into from
Sep 10, 2020

Conversation

lznamens
Copy link
Contributor

@lznamens lznamens commented Sep 7, 2020

Optimization of the kernel "fully_connected_gpu_MMAD":

  • increasing number of OpenCL work-items using partial sums with SLM for cases with low GPU threads loading (such as alexnet);
  • manual unrolling for decreasing number of instruction mixings from compiler.

JIRA: 32370

@lznamens lznamens added the category: GPU OpenVINO GPU plugin label Sep 7, 2020
@lznamens lznamens added this to the 2021.1 milestone Sep 7, 2020
@lznamens lznamens requested review from a team as code owners September 7, 2020 17:33
@lznamens lznamens force-pushed the feature/iznamenskiy/fully_connected_mmad_i8_improvements branch from 3075f12 to 4514b29 Compare September 8, 2020 09:17
@lznamens lznamens force-pushed the feature/iznamenskiy/fully_connected_mmad_i8_improvements branch from 4514b29 to 5d7bf81 Compare September 8, 2020 09:25
@vladimir-paramuzov vladimir-paramuzov merged commit 3797a28 into openvinotoolkit:master Sep 10, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@sshlyapn sshlyapn sshlyapn left review comments

@vladimir-paramuzov vladimir-paramuzov vladimir-paramuzov approved these changes

Assignees
No one assigned
Labels
category: GPU OpenVINO GPU plugin
Projects
None yet
Milestone
2021.1
Development

Successfully merging this pull request may close these issues.
3 participants
@lznamens @sshlyapn @vladimir-paramuzov