[gpuCI] Forward-merge branch-21.10 to branch-21.12 [skip gpuci] #333
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Removes `-g` from the compile commands generated by distutils to compile Cython files. This will make our container images, conda packages, and python wheels smaller.
…treams (#329) Comes from this PR rapidsai/rmm#873 Authors: - Divye Gala (https://github.com/divyegala) Approvers: - Chuck Hastings (https://github.com/ChuckHastings) - Dante Gama Dessavre (https://github.com/dantegd) URL: #329
…y gmem to store intermediate distances (#324) benchmarking with cuml python interface kNN datasets (it claims to generate gaussian distribution) tried till 200k x 128 database/query vectors. found some different behavior on my small GPU GP107 vs on GA102(Tesla A40) on GP107 fused L2 kNN is slower on larger datasets on GA102 fused L2 kNN is always faster like approx **1.15x-1.5x** for all datasets I tried (except 200k x 128). I will also have L2 expanded version of fused L2 kNN in a separate PR due to which on larger dimension like > 128 distance computation from fused L2 kNN won't become bottleneck. There is scope to optimize the distance computation in fused L2 kNN as there is no usage of vectorized LDG/STS in it. Overall it looks that fused L2 kNN is better on GPUs with decent compute power but not on small old GPUs like GP107. & benchmarking with cuml cpp kNN regression tests the performance is On A30 (GA100) , For NN == 64, resultant Dist matrix 1M x 1M, Fused L2 kNN = 11550ms FAISS kNN = 23933 ms. **Overall 2.07x faster** And for NN == 32, it is **1.43x faster** runtimes for NN == 32, Fused L2 kNN = 11198ms FAISS kNN = 16124ms Authors: - Mahesh Doijade (https://github.com/mdoijade) Approvers: - Chuck Hastings (https://github.com/ChuckHastings) - Corey J. Nolet (https://github.com/cjnolet) URL: #324
When we make a new raft version, we need to also bump the rapids-cmake version at the same time. Otherwise we will get the previous releases dependencies by mistake. Authors: - Robert Maynard (https://github.com/robertmaynard) Approvers: - AJ Schmidt (https://github.com/ajschmidt8) - Dante Gama Dessavre (https://github.com/dantegd) URL: #331
|
SUCCESS - forward-merge complete. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Forward-merge triggered by push to
branch-21.10that creates a PR to keepbranch-21.12up-to-date. If this PR is unable to be immediately merged due to conflicts, it will remain open for the team to manually merge.