-
Notifications
You must be signed in to change notification settings - Fork 87
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
Prettify README #1504
Prettify README #1504
Conversation
4791994
to
7dfe9de
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## develop #1504 +/- ##
===========================================
- Coverage 89.37% 89.33% -0.05%
===========================================
Files 696 688 -8
Lines 56944 56557 -387
===========================================
- Hits 50895 50525 -370
+ Misses 6049 6032 -17 ☔ View full report in Codecov by Sentry. |
21ed856
to
e67bc0b
Compare
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.
nit on the badge margin but it is not critical.
@@ -73,7 +75,7 @@ The Ginkgo DPC++(SYCL) module has the following __additional__ requirements: | |||
|
|||
The Ginkgo MPI module has the following __additional__ requirements: | |||
|
|||
* MPI 3.1+, ideally with GPUDirect support for best performance | |||
* MPI 3.1+, ideally GPU-Aware, for best performance |
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.
I think GPU-Aware is different from GPUDirect.
GPU-Aware only recognizes whether the memory is on GPU such that we do not need to move the data to host.
GPU direct is more relevant for performance
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.
No, GPUDirect is a NVIDIA related terminology that is unrelated to MPI. The correct term for MPI implementation is GPU-Aware is used across all MPI implementations. See for example paper from the MVAPICH2 team
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.
GPU-Aware means that the MPI implementation is able to accept GPU pointers, it does not say whether they actually use direct connections between GPUs. That is what GPUDirect refers to on the NVIDIA side, I'm not sure if there is an equivalent for AMD or Intel.
e67bc0b
to
81e90cd
Compare
Co-authored-by: Marcel Koch <[email protected]> Co-authored-by: Yu-Hsiang Tsai <[email protected]> Co-authored-by: Gregor Olenik <[email protected]>
81e90cd
to
f1026b9
Compare
Error: PR already merged! |
Quality Gate passedKudos, no new issues were introduced! 0 New issues |
This PR makes the README a bit more pretty and adds a TOC, which I think should improve readability.
See how it looks by going to the branch README file