perf: avoid redundant dp.compute() across MPI ranks#7524
Open
curry-sthuc wants to merge 2 commits into
Open
Conversation
added 2 commits
June 26, 2026 16:23
Only rank 0 calls dp.compute() and broadcasts results to all ranks via MPI_Bcast, avoiding N-fold redundant computation and CPU contention.
Only rank 0 needs coord and cell vectors for dp.compute(). Non-rank-0 ranks receive results via MPI_Bcast and never use these vectors.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
In
ESolver_DP::runner(), every MPI rank callsdp.compute()independently with identical input, producing the same energy, forces, and virial. With N ranks, this results in N-fold redundant computation, and multiple concurrent deepmd inference calls cause CPU contention.Solution
Only rank 0 calls
dp.compute(), then broadcasts results viaMPI_Bcast. Implemented with#ifdef __MPIguards — serial builds are unaffected.Performance (864 Al atoms, 100 MD steps)
Checklist