Add collectives configurations for different slice size #81
+156
−16
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Outdated suggestions cannot be applied.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.
Summary
This PR introduces configurations for various Ironwood topologies (ranging from
2x2x1to4x4x4) and updates the time extraction logic for accurate benchmarking.Key Changes
all-gather,psum(all-reduce), andall-to-all) for multiple topologies inIronwood/configs/collectives/.Ironwood/src/benchmark_utils.pyto correctly accumulate execution time for splitpsumoperations.Reason for Logic Change
Previously, a single JAX
psumoperation generated two distinct HLOs (psumandffi_call). The old logic treated these as two separate collective operations (e.g., ifnum_runs=5, it extracted 10 events).This caused a
ZeroDivisionErrorbecause theffi_callcompletes in near 0ms. From a collectives perspective, these two HLOs represent a single logical step, so their execution times are now aggregated to prevent errors and ensure accurate timing.