Fix performance regression in UnitarySynthesis #13375
Merged
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
This commit fixes a performance regression that was introduced in PR Qiskit#13141. When the pass is looking up the preferred synthesis direction for a unitary based on the connectvity constraints the connectivity was being provided as a PyList. To look up the edge in connectivity set this meant we needed to iterate over the list and then create a set that rust could lookup if it contains an edge or it's reverse. This has significant overhead because its iterating via python and also iterating per decomposition. This commit addresses this by changing the input type to be a HashSet from Python so Pyo3 will convert a pyset directly to a HashSet once at call time and that's used by reference for lookups directly instead of needing to iterate over the list each time.
mtreinish
added
performance
Changelog: None
mtreinish
requested review from
alexanderivrii,
ShellyGarion and
a team
as code owners
|
One or more of the following people are relevant to this code:
|
Pull Request Test Coverage Report for Build 11575706080Details
💛 - Coveralls |
If we're using the default plugin the execution will happen in rust and we don't need to build the plugin data views that are defined in the plugin interface. Profiling the benchpress test using the hamlib hamiltonian: ham_graph-2D-grid-nonpbc-qubitnodes_Lx-5_Ly-186_h-0.5-all-to-all that caught this originally regression was showing an inordinate amount of time being spent in the construction of `_build_gate_lengths_by_qubit` and `_build_gate_errors_by_qubit` which isn't being used because this happens internally in rust now. To mitigate this overhead this commit migrates the sections computing these values to the code branch that uses them and not the default plugin branch that uses rust.
mtreinish
commented
Oct 29, 2024
| @@ -268,7 +268,7 @@ fn py_run_main_loop( | |||
| new_ids, | |||
| min_qubits, | |||
| target, | |||
| coupling_edges, | |||
| coupling_edges.clone(), | |||
There was a problem hiding this comment.
We can avoid this clone if we refactor the functions a bit to split the python interface out and use a reference for the recursion. Not sure how important this is because this only comes up during control flow processing and we're already slower there because of the python usage and circuit to dag conversion. Maybe something to fix for when we make a dag native control flow op.
There was a problem hiding this comment.
I think it would make sense to revisit this when we get to have rust-native control flow op.
ElePT
approved these changes
Oct 29, 2024
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.
Summary
This commit fixes a performance regression that was introduced in PR #13141. When the pass is looking up the preferred synthesis direction for a unitary based on the connectvity constraints the connectivity was being provided as a PyList. To look up the edge in connectivity set this meant we needed to iterate over the list and then create a set that rust could lookup if it contains an edge or it's reverse. This has significant overhead because its iterating via python and also iterating per decomposition. This commit addresses this by changing the input type to be a HashSet from Python so Pyo3 will convert a pyset directly to a HashSet once at call time and that's used by reference for lookups directly instead of needing to iterate over the list each time.
Details and comments