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

Implement more asynchronous dependency handling on GPU #2963

Merged
merged 3 commits into from Jun 22, 2021
Merged

Implement more asynchronous dependency handling on GPU #2963

merged 3 commits into from Jun 22, 2021

Commits on Jun 9, 2021

  1. Replace CPU synchronization with cudaStreamWaitEvent to asynchronousl… 

    …y handle data dependencies of GPU operations.

Return cudaEvent on GPU op completion to framework to enable asynchronous callbacks (enabled with HOROVOD_ENABLE_ASYNC_COMPLETION).

Signed-off-by: Josh Romero <joshr@nvidia.com>
    @romerojosh
    romerojosh committed Jun 9, 2021
    Configuration menu
    Copy the full SHA
    73e9077 View commit details
    Browse the repository at this point in the history

  2. Add missing HAVE_GPU guards. 

    Signed-off-by: Josh Romero <joshr@nvidia.com>
    @romerojosh
    romerojosh committed Jun 9, 2021
    Configuration menu
    Copy the full SHA
    104485b View commit details
    Browse the repository at this point in the history

  3. CPU compilation fix. 

    Signed-off-by: Josh Romero <joshr@nvidia.com>
    @romerojosh
    romerojosh committed Jun 9, 2021
    Configuration menu
    Copy the full SHA
    eb09473 View commit details
    Browse the repository at this point in the history