[WIP] Add new multithreaded TwoQubitPeepholeOptimization pass #13419

mtreinish · 2024-11-10T16:12:54Z

Summary

This commit adds a new transpiler pass for physical optimization,
TwoQubitPeepholeOptimization. This replaces the use of Collect2qBlocks,
ConsolidateBlocks, and UnitarySynthesis in the optimization stage for
a default pass manager setup. The pass logically works the same way
where it analyzes the dag to get a list of 2q runs, calculates the matrix
of each run, and then synthesizes the matrix and substitutes it inplace.
The distinction this pass makes though is it does this all in a single
pass and also parallelizes the matrix calculation and synthesis steps
because there is no data dependency there.

This new pass is not meant to fully replace the Collect2qBlocks,
ConsolidateBlocks, or UnitarySynthesis passes as those also run in
contexts where we don't have a physical circuit. This is meant instead
to replace their usage in the optimization stage only. Accordingly this
new pass also changes the logic on how we select the synthesis to use
and when to make a substitution. Previously this logic was primarily done
via the ConsolidateBlocks pass by only consolidating to a UnitaryGate if
the number of basis gates needed based on the weyl chamber coordinates
was less than the number of 2q gates in the block (see #11659 for
discussion on this). Since this new pass skips the explicit
consolidation stage we go ahead and try all the available synthesizers

Right now this commit has a number of limitations, the largest are:

Only supports the target
It doesn't support the XX decomposer because it's not in rust (the TwoQubitBasisDecomposer and TwoQubitControlledUDecomposer are used)

This pass doesn't support using the unitary synthesis plugin interface, since
it's optimized to use Qiskit's built-in two qubit synthesis routines written in
Rust. The existing combination of ConsolidateBlocks and UnitarySynthesis
should be used instead if the plugin interface is necessary.

Details and comments

Fixes #12007
Fixes #11659

TODO:

Rebase after Use OnceLock instead of OnceCell #13410 merges
Add tests
Add documentation
Benchmarking and performance tuning
Handle running serially when in multiprocessing context
Release note

coveralls · 2024-11-10T16:37:45Z

Pull Request Test Coverage Report for Build 14525825123

Warning: This coverage report may be inaccurate.

This pull request's base commit is no longer the HEAD commit of its target branch. This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.

For more information on this, see Tracking coverage changes with pull request builds.
To avoid this issue with future PRs, see these Recommended CI Configurations.
For a quick fix, rebase this PR at GitHub. Your next report should be accurate.

Details

410 of 441 (92.97%) changed or added relevant lines in 11 files are covered.
393 unchanged lines in 4 files lost coverage.
Overall coverage increased (+0.07%) to 88.3%

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
crates/accelerate/src/two_qubit_decompose.rs	13	14	92.86%
crates/accelerate/src/unitary_synthesis.rs	2	3	66.67%
crates/accelerate/src/two_qubit_peephole.rs	359	388	92.53%

Files with Coverage Reduction	New Missed Lines	%
crates/qasm2/src/lex.rs	2	93.23%
crates/accelerate/src/unitary_synthesis.rs	31	94.69%
crates/accelerate/src/basis/basis_translator/mod.rs	37	89.48%
crates/circuit/src/dag_circuit.rs	323	86.88%

Totals
Change from base Build 14525370807:	0.07%
Covered Lines:	74214
Relevant Lines:	84048

💛 - Coveralls

This commit adds a new transpiler pass for physical optimization, TwoQubitPeepholeOptimization. This replaces the use of Collect2qBlocks, ConsolidateBlocks, and UnitarySynthesis in the optimization stage for a default pass manager setup. The pass logically works the same way where it analyzes the dag to get a list of 2q runs, calculates the matrix of each run, and then synthesizes the matrix and substitutes it inplace. The distinction this pass makes though is it does this all in a single pass and also parallelizes the matrix calculation and synthesis steps because there is no data dependency there. This new pass is not meant to fully replace the Collect2qBlocks, ConsolidateBlocks, or UnitarySynthesis passes as those also run in contexts where we don't have a physical circuit. This is meant instead to replace their usage in the optimization stage only. Accordingly this new pass also changes the logic on how we select the synthesis to use and when to make a substituion. Previously this logic was primarily done via the ConsolidateBlocks pass by only consolidating to a UnitaryGate if the number of basis gates needed based on the weyl chamber coordinates was less than the number of 2q gates in the block (see Qiskit#11659 for discussion on this). Since this new pass skips the explicit consolidation stage we go ahead and try all the available synthesizers Right now this commit has a number of limitations, the largest are: - Only supports the target - It doesn't support any synthesizers besides the TwoQubitBasisDecomposer, because it's the only one in rust currently. For plugin handling I left the logic as running the three pass series, but I'm not sure this is the behavior we want. We could say keep the synthesis plugins for `UnitarySynthesis` only and then rely on our built-in methods for physical optimiztion only. But this also seems less than ideal because the plugin mechanism is how we support synthesizing to custom basis gates, and also more advanced approximate synthesis methods. Both of those are things we need to do as part of the synthesis here. Additionally, this is currently missing tests and documentation and while running it manually "works" as in it returns a circuit that looks valid, I've not done any validation yet. This also likely will need several rounds of performance optimization and tuning. t this point this is just a rough proof of concept and will need a lof refinement along with larger changes to Qiskit's rust code before this is ready to merge. Fixes Qiskit#12007 Fixes Qiskit#11659

…rallel-pass

Since Qiskit#13139 merged we have another two qubit decomposer available to run in rust, the TwoQubitControlledUDecomposer. This commit updates the new TwoQubitPeepholeOptimization to call this decomposer if the target supports appropriate 2q gates.

Clippy is correctly warning that the size difference between the two decomposer types in the TwoQubitDecomposer enumese two types is large. TwoQubitBasisDecomposer is 1640 bytes and TwoQubitControlledUDecomposer is only 24 bytes. This means each element of ControlledU is wasting > 1600 bytes. However, in this case that is acceptable in order to avoid a layer of pointer indirection as these are stored temporarily in a vec inside a thread to decompose a unitary. A trait would be more natural for this to define a common interface between all the two qubit decomposers but since we keep them instantiated for each edge in a Vec they need to be sized and doing something like `Box<dyn TwoQubitDecomposer>` (assuming a trait `TwoQubitDecomposer` instead of a enum) to get around this would have additional runtime overhead. This is also considering that TwoQubitControlledUDecomposer has far less likelihood in practice as it only works with some targets that have RZZ, RXX, RYY, or RZX gates on an edge which is less common.

…rallel-pass

Also don't run scoring more than needed.

ShellyGarion · 2025-01-23T07:02:09Z

Copy here the comment of @t-imamichi #13568 (comment)
and my reply: #13568 (comment)

I think this closes #13428. How about adding a test case of consecutive RZZ (RXX, and RYY) gates?

We should make sure that after PR #13568 and this PR will be merged, we can efficiently transpile circuits into basis fractional RZZ gates .

…rallel-pass

mtreinish · 2025-01-26T14:09:40Z

I added support for using the ControlledUDecomposer to the new pass back in early December with this commit: 746758f although looking at that now with fresh eyes I need to check that the gate is continuous in the target, right now it only looks at the supported gate types.

…tinuous

The priority for the two qubit peephole pass should be decreasing the 2q gate count. The error rate heuristic should only matter if the 2q counts are the same. This commit flips the heuristic to first check the 2q gate count so the first priority is reducing the 2q gate count.

jlapeyre · 2025-02-20T13:26:02Z

qiskit/transpiler/passes/optimization/two_qubit_peephole.py

@@ -0,0 +1,102 @@
+# This code is part of Qiskit.
+#
+# (C) Copyright IBM 2017, 2024.


Suggested change

# (C) Copyright IBM 2017, 2024.

# (C) Copyright IBM 2017, 2025.

Seems like the copyright years have not often been updated in source files. But since this is a new file, and a new year...

I did write the pass originally in 2024. I've been working on this on and off since November.

jlapeyre · 2025-02-20T13:26:16Z

crates/accelerate/src/two_qubit_peephole.rs

@@ -0,0 +1,533 @@
+// This code is part of Qiskit.
+//
+// (C) Copyright IBM 2024


Suggested change

// (C) Copyright IBM 2024

// (C) Copyright IBM 2025

ShellyGarion · 2025-03-11T13:07:32Z

Copy here this #13428 (comment) - is this example being solved by this PR?

One case that has not been solved yet is the following:

the circuit has one CZ gate and one RZZ gate
the target has basis gates containing 'cz' and 'rzz'.

In this case, the circuit would not get consolidated and re-synthesized into a single RZZ gate.
I'm not sure that I have a quick fix for this, but hope that it would be solved in #13419.

       basis_gates = ["rzz", "rx", "rz", "cz"]
        qc = QuantumCircuit(2)
        qc.rzz(0.1, 0, 1)
        qc.cz(0, 1)
        consolidate_pass = ConsolidateBlocks(basis_gates=basis_gates)
        res = consolidate_pass(qc)
        pm = generate_preset_pass_manager(optimization_level=2, basis_gates=basis_gates)
        tqc = pm.run(qc)
        print (res)
        print (tqc)

outputs:


q_0: ─■─────────■─
      │ZZ(0.1)  │ 
q_1: ─■─────────■─
                  
                  
q_0: ─■─────────■─
      │ZZ(0.1)  │ 
q_1: ─■─────────■─

ShellyGarion · 2025-03-11T13:17:08Z

Is it possible to try to synthesize both U and U^(-1) to check which one produces a better synthesis?
We may be able to reduce some 1-qubit gates this way (as one can always invert the gates in the synthesized circuit).
See the following example:

        qc = QuantumCircuit(2)
        qc.rzz(0.1, 0, 1)
        qc.rzz(0.2, 0, 1)
        pm = generate_preset_pass_manager(optimization_level=2, basis_gates=["rzz", "rx", "rz"])
        tqc = pm.run(qc)
        print (tqc)

produces the circuit:

global phase: π
                                  
q_0: ──────────■──────────────────
     ┌───────┐ │ZZ(-0.3) ┌───────┐
q_1: ┤ Rx(π) ├─■─────────┤ Rx(π) ├
     └───────┘           └───────┘

while its inverse:

        tqc = pm.run(qc.inverse())
        print (tqc)

produces the circuit:

q_0: ─■─────────
      │ZZ(-0.3) 
q_1: ─■─────────

…rallel-pass

This commit removes the unitary synthesis plugin mechanism from the pass. This was a layer violation to support this when the pass logic doesn't actually support using the plugin interface. It is easier and more clear that if the plugin interface usage is desired to handle that in the pass manager construction rather than have this pass internally build a pass manager and execute other passes to emulate behavior it doesn't have.

…rallel-pass

There were two issues identified by the testing which required fixing and adjusting the tests based on limitations in the pass. The first issue was the parameters for the target gate was not handled correctly. In the case of using the Controlled U decomposer we were not passing the computed parameter value correctly to the output circuit and instead the ParameterExpression from the target was being used. Then in the case of controlled gates (not supercontrolled) that had a fixed angle that are normally intended for the xx decomposer were incorrectly being passed to the TwoQubitBasisDecomposer which can't work with them. This was resulting in invalid circuit outputs. The use of the TwoQubitBasisDecomposer is now correctly filtering to only be run with supercontrolled gates. The tests were adjusted for this limitation because they were mostly copied from the UnitarySynthesis tests which supports xx decomposer.

…rallel-pass

t-imamichi · 2025-04-18T02:16:12Z

test/python/transpiler/test_two_qubit_peephole.py

+    UGate,
+    ZGate,
+    RYYGate,
+    RZZGate,


According to the error log, RZZGate is not used.

Suggested change

RZZGate,

ShellyGarion · 2025-04-18T05:11:04Z

I've suggested some further tests here: mtreinish#31

…rallel-pass

The typing for some of the new methods on the DAGCircuitBuilder where a bit too strict and requried the caller to do more work than was necessary. This commit loosens the typing to make it a bit more ergonomic and straightforward to use. It also more closely matches the DAGCircuit methods the builder struct is mirroring. Right now the only signature difference is qubits and clbits are wrapped in an Option while on DAGCircuit it's not. This commit doesn't change this difference, although there really isn't a reason to make this distinction and both methods could have the same signature.

This commit updates some of the pass's logic and handles some bugs that were found during development. It also adds a couple new test cases, one of which is failing. There still seem to be bugs in the case we need to use a reverse edge in the target.

mtreinish added performance Changelog: New Feature Include in the "Added" section of the changelog Rust This PR or issue is related to Rust code in the repository mod: transpiler Issues and PRs related to Transpiler labels Nov 10, 2024

mtreinish added this to the 2.0.0 milestone Nov 10, 2024

This was referenced Nov 11, 2024

Use OnceLock instead of OnceCell #13410

Merged

Suboptimal optimization result of consecutive RZZ gates with optimization level 2 and 3 #13431

Open

mtreinish force-pushed the two-qubit-peephole-parallel-pass branch from ad06d1a to 4d160bc Compare November 14, 2024 12:34

mtreinish mentioned this pull request Nov 18, 2024

Fix post-oxidization change in ConsolidateBlocks behavior #13450

Merged

mtreinish added 8 commits December 1, 2024 02:05

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

decee9a

…rallel-pass

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

4d4df68

…rallel-pass

Embed 2q gate count into score as tie breaker

cb6b70f

Also don't run scoring more than needed.

Release GIL during parallel portion

f06a070

Merge branch 'main' into two-qubit-peephole-parallel-pass

90b16e8

Fix lint

a175ee8

ShellyGarion mentioned this pull request Jan 23, 2025

Add 2q fractional gates to the UnitarySynthesis transpiler pass #13568

Merged

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

af0c144

…rallel-pass

mtreinish added 4 commits January 26, 2025 09:28

Update ControlledUDecomposer to ensure we only run if the gate is con…

79a46c5

…tinuous

Add reversed synthesis for two qubit basis decomposer

839b4c9

Fix handling of single direction gates

d9399a6

Fix import cycle

b4c4360

ShellyGarion self-assigned this Feb 3, 2025

Merge branch 'main' into two-qubit-peephole-parallel-pass

aefdc90

1ucian0 assigned mtreinish Feb 6, 2025

mtreinish added 5 commits February 19, 2025 17:48

Add docstring to new pass

55b05c0

Add release note

7756d1e

Run serially in multiprocessing context

6a01332

Fix cache build

d83562b

jlapeyre reviewed Feb 20, 2025

View reviewed changes

mtreinish modified the milestones: 2.0.0, 2.1.0 Feb 21, 2025

Add tests

5f73b93

ShellyGarion mentioned this pull request Mar 11, 2025

Optimize consecutive RZZ gates into a single RZZ gate #13428

Closed

mtreinish added 10 commits April 11, 2025 15:53

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

5e6c4cb

…rallel-pass

Rebase updates

8c7e67c

Adjust tests

f41e855

Clean-up test lint

3cdee4f

Fix oversight in test code

89931a4

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

19368bb

…rallel-pass

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

02785e1

…rallel-pass

Update target usage for recent changes

c7ba671

t-imamichi reviewed Apr 18, 2025

View reviewed changes

mtreinish added 5 commits April 18, 2025 06:23

Merge remote-tracking branch 'origin/main' into two-qubit-peephole-pa…

089d3b6

…rallel-pass

Use dag builder api

bfdfe46

Fix error handling for target lookup

f72df7d

Fix some bugs and expand tests

bb31eb5

This commit updates some of the pass's logic and handles some bugs that were found during development. It also adds a couple new test cases, one of which is failing. There still seem to be bugs in the case we need to use a reverse edge in the target.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[WIP] Add new multithreaded TwoQubitPeepholeOptimization pass #13419

[WIP] Add new multithreaded TwoQubitPeepholeOptimization pass #13419

mtreinish commented Nov 10, 2024 •

edited

Loading

coveralls commented Nov 10, 2024 •

edited

Loading

ShellyGarion commented Jan 23, 2025

mtreinish commented Jan 26, 2025

jlapeyre Feb 20, 2025

mtreinish Feb 20, 2025

jlapeyre Feb 20, 2025

ShellyGarion commented Mar 11, 2025

ShellyGarion commented Mar 11, 2025 •

edited

Loading

t-imamichi Apr 18, 2025 •

edited

Loading

ShellyGarion commented Apr 18, 2025

	# (C) Copyright IBM 2017, 2024.
	# (C) Copyright IBM 2017, 2025.

[WIP] Add new multithreaded TwoQubitPeepholeOptimization pass #13419

Are you sure you want to change the base?

[WIP] Add new multithreaded TwoQubitPeepholeOptimization pass #13419

Conversation

mtreinish commented Nov 10, 2024 • edited Loading

Summary

Details and comments

coveralls commented Nov 10, 2024 • edited Loading

Pull Request Test Coverage Report for Build 14525825123

Warning: This coverage report may be inaccurate.

Details

💛 - Coveralls

ShellyGarion commented Jan 23, 2025

mtreinish commented Jan 26, 2025

jlapeyre Feb 20, 2025

Choose a reason for hiding this comment

mtreinish Feb 20, 2025

Choose a reason for hiding this comment

jlapeyre Feb 20, 2025

Choose a reason for hiding this comment

ShellyGarion commented Mar 11, 2025

ShellyGarion commented Mar 11, 2025 • edited Loading

t-imamichi Apr 18, 2025 • edited Loading

Choose a reason for hiding this comment

ShellyGarion commented Apr 18, 2025

mtreinish commented Nov 10, 2024 •

edited

Loading

coveralls commented Nov 10, 2024 •

edited

Loading

ShellyGarion commented Mar 11, 2025 •

edited

Loading

t-imamichi Apr 18, 2025 •

edited

Loading