Add a script to run PCA and clustering of our benchmarks#317
Open
fitzgen wants to merge 1 commit into
Open
Conversation
The methodology is based on ["A Workload Characterization of the SPEC CPU2017 Benchmark Suite" by Limaye and Adegbija](https://tosiron.com/papers/2018/SPEC2017_ISPASS18.pdf). Each metric is standardized (centered to mean 0, scaled to unit variance) and PCA is run on the resulting correlation matrix so that metrics measured on different scales contribute comparably. Benchmarks are then clustered by the Euclidean distance between their principal-component scores, as in the paper. Finally, we recommend a subset of the suite. Each cluster is represented by its cheapest member (the benchmark that executes the fewest dynamic wasm instructions). Sweeping the number of clusters traces a Pareto trade-off between clustering error (SSE) and the cost of running the subset (its total dynamic instructions); the knee of that curve is the Pareto-optimal cluster size. Fixes bytecodealliance#98
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The methodology is based on "A Workload Characterization of the SPEC CPU2017 Benchmark Suite" by Limaye and
Adegbija.
Each metric is standardized (centered to mean 0, scaled to unit variance) and PCA is run on the resulting correlation matrix so that metrics measured on different scales contribute comparably. Benchmarks are then clustered by the Euclidean distance between their principal-component scores, as in the paper.
Finally, we recommend a subset of the suite. Each cluster is represented by its cheapest member (the benchmark that executes the fewest dynamic wasm instructions). Sweeping the number of clusters traces a Pareto trade-off between clustering error (SSE) and the cost of running the subset (its total dynamic instructions); the knee of that curve is the Pareto-optimal cluster size.
Fixes #98