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.

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

Already on GitHub? Sign in to your account

Add support for IPC features for PrivateUse devices #124902

Open
mengpenghui opened this issue Apr 25, 2024 · 0 comments 路 May be fixed by #125122
Open

Add support for IPC features for PrivateUse devices #124902

mengpenghui opened this issue Apr 25, 2024 · 0 comments 路 May be fixed by #125122
Labels
module: PrivateUse1 private use oncall: distributed Add this issue/PR to distributed oncall triage queue

Comments

@mengpenghui
Copy link

mengpenghui commented Apr 25, 2024

馃殌 The feature, motivation and pitch

Currently, PyTorch only supports inter-process communication (IPC) on CUDA devices. I'm looking into doing IPC on 3rd party devices and hope some abstract code can be added to the PyTorch community.
So I want to ask if this feature is considered

Alternatives

I plan to provide underlying support through PrivateUse1Hooks, and provide PrivateUse1 device support for the share_device, _new_shared_device, and _release_ipc_counter_device methods in the python Storage class.

Additional context

No response

cc @mrshenli @pritamdamania87 @zhaojuanmao @satgera @rohan-varma @gqchen @aazzolini @osalpekar @jiayisuse @H-Huang @kwen2501 @awgu @penguinwu @fegin @XilunWu @wanchaol @fduwjj @wz337 @tianyu-l @wconstab @yf225 @chauhang @d4l3k

@soulitzer soulitzer added oncall: distributed Add this issue/PR to distributed oncall triage queue module: PrivateUse1 private use labels Apr 25, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue Apr 28, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue Apr 28, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue Apr 29, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue Apr 29, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue May 6, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue May 6, 2024
mengpenghui added a commit to mengpenghui/pytorch that referenced this issue May 6, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
module: PrivateUse1 private use oncall: distributed Add this issue/PR to distributed oncall triage queue
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants