-
Notifications
You must be signed in to change notification settings - Fork 21
Open
Description
Tell us the versions of your environment (ensure you are running the latest):
- forest benchmarking version: 0.8.0 (pre-release branch)
- pyQuil version: 3.0.0
- qvm version: latest
- quilc version: 1.23.0
The documentation for randomized benchmarking mentions capabilities for RB, simultaneous RB, and interleaved RB. All experiments are generated via generate_rb_experiments(). Simultaneous RB takes in multiple groups; interleaved RB takes in an interleaved gate.
As it turns out, you cannot do simultaneous, interleaved RB. If you try, e.g. providing an interleaved gate program with several 2Q gates and matching groups, you will get an error. Such a use case it a little nonsensical given the restriction for serial 2Q execution on QPU.
To reproduce:
benchmarker = BenchmarkConnection() # from pyquil v3
program = Program()
program += CZ(30, 31)
program += CZ(32, 33)
generate_rb_experiments(benchmarker, [(30, 31), (32, 33)], [2, 10], interleaved_gate=program)
Recommend a consistency check be added to fail gracefully.
Metadata
Metadata
Assignees
Labels
No labels