-
Notifications
You must be signed in to change notification settings - Fork 3k
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’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Performance] Use segment operators for graph readout. #2361
Conversation
} else { | ||
cuda::AtomicAdd(out_buf, val); | ||
} | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We should refactor this part.
*/ | ||
template <typename IdType, typename DType, | ||
typename ReduceOp> | ||
__global__ void SegmentReduceKernel( |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
SegmentReduceKernel -> SegmentReduceCmpKernel
Also, poor docstring
Also missing C++ tests |
* upd * upd * update * upd * upd * upd * fix * lint * lint * pylint * doc
Will fix the documentation issue soon, for other concerns:
|
Description
Previously we created a pseudo graph and applied SpMM to get graph readout results, and introduced graph creation overhead. Thus PR implements segment operators to lower this (and corresponding coo/csr conversion) overhead.
Checklist
Please feel free to remove inapplicable items for your PR.
or have been fixed to be compatible with this change