Benchmarking protocol for quantum gates
Systems and methods for performing iFRB benchmarking are disclosed. Exemplary systems can include a quantum component and a classical component. The classical component can include at least one processor and at least one non-transitory computer-readable medium. The non-transitory computer-readable medium can contain instructions that, when executed by the at least one processor, cause the classical component to perform operations. The operations can include generating a gate sequence including test gates interleaved with gate subsequences equivalent to Haar random two-qubit gates. Each gate subsequence can include at least one SQiSW gate. The operations can further include generating a recovery gate based on the gate sequence. The operations can further include providing commands applying the gate sequence and the recovery gate to the quantum component and obtaining an output from the quantum component.
1. A system for performing Interleaved Fully Randomized Benchmarking (iFRB), comprising:
a quantum component; and
a classical component, the classical component including at least one processor, and at least one non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the classical component to perform operations comprising:
generating a gate sequence including test gates interleaved with gate subsequences equivalent to Haar random two-qubit gates, each gate subsequence including at least one SQiSW gate;
generating a recovery gate based on the gate sequence; and
providing commands applying the gate sequence and the recovery gate to the quantum component and obtaining an output from the quantum component.
2. The system of claim 1 , wherein:
generating one of the gate subsequences comprises generating a Haar random two-qubit gate and decomposing, using two Cartan decompositions, the Haar random two-qubit gate into the one of the gate subsequences.
3. The system of claim 1 , wherein:
generating one of the gate subsequences comprises:
sampling a point in a Weyl chamber; and
generating the one of the gate subsequences based on the sampled point.
4. The system of claim 3 , wherein:
generating the one of the gate subsequences further comprises sampling four Haar-random single-qubit gates; and
the one of the gate subsequences is generated using the sampled four Haar-random single-qubit gates.
5. The system of claim 1 , wherein:
generating one of the gate subsequences comprises generating three interaction coefficients;
when the three interaction coefficients satisfy a Weyl condition, the one of the gate subsequences includes two SQiSW gates; and
when the three interaction coefficients do not satisfy the Weyl condition, the one of the gate subsequences includes three SQiSW gates.
6. The system of claim 5 , wherein:
generating the one of the gate subsequences further comprises:
determining interleaving single qubit gates based on the three interaction coefficients.
7. The system of claim 6 , wherein:
generating the one of the gate subsequences further comprises:
sampling four Haar-random single-qubit gates; and
generating the one of the gate subsequences using the interleaving single qubit gates and the four sampled Haar-random single-qubit gates.
8. The system of claim 5 , wherein:
generating one of the gate subsequences comprises:
generating three canonicalized interaction coefficients and two first single-qubit gates using the three interaction coefficients;
generating two second single-qubit gates by performing a Cartan decomposition on a gate corresponding to the three canonicalized interaction coefficients;
determining second interleaving single-qubit gates based on the three canonicalized interaction coefficients; and
generating the one of the gate subsequences using the second interleaving single-qubit gates, the two first single-qubit gates, and the two second single-qubit gates.
9. The system of claim 1 , wherein:
the test gates comprise SQiSW gates.
10. A non-transitory computer-readable medium containing instructions that, when executed by at least one processor, cause a system to perform operations for performing iFRB benchmarking, comprising:
generating a gate sequence including test gates interleaved with gate subsequences equivalent to Haar random two-qubit gates, each gate subsequence including at least one SQiSW gate;
generating a recovery gate based on the gate sequence; and
providing commands applying the gate sequence and the recovery gate to a quantum component and obtaining an output from the quantum component.
11. The non-transitory computer-readable medium of claim 10 , wherein:
generating one of the gate subsequences comprises generating a Haar random two-qubit gate and decomposing, using two Cartan decompositions, the Haar random two-qubit gate into the one of the gate subsequences.
12. The non-transitory computer-readable medium of claim 10 , wherein:
generating one of the gate subsequences comprises:
sampling a point in a Weyl chamber; and
generating the one of the gate subsequences based on the sampled point.
13. The non-transitory computer-readable medium of claim 10 , wherein:
generating one of the gate subsequences comprises generating three interaction coefficients;
when the three interaction coefficients satisfy a Weyl condition the one of the gate subsequences includes two SQiSW gates; and
when the three interaction coefficients do not satisfy the Weyl condition the one of the gate subsequences includes three SQiSW gates.
14. The non-transitory computer-readable medium of claim 13 , wherein:
generating the one of the gate subsequences further comprises:
determining interleaving single qubit gates based on the three interaction coefficients;
sampling four Haar-random single-qubit gates; and
generating the one of the gate subsequences using, in part, the interleaving single qubit gates and the four sampled Haar-random single-qubit gates.
15. The non-transitory computer-readable medium of claim 13 , wherein:
generating one of the gate subsequences comprises:
generating three canonicalized interaction coefficients and two first single-qubit gates using the three interaction coefficients;
generating two second single-qubit gates by performing a Cartan decomposition on a gate corresponding to the three canonicalized interaction coefficients;
determining second interleaving single-qubit gates based on the three canonicalized interaction coefficients; and
generating the one of the gate subsequences using, in part, the second interleaving single-qubit gates, the two first single-qubit gates, and the two second single-qubit gates.
16. A method of performing iFRB benchmarking, comprising:
generating a gate sequence including test gates interleaved with gate subsequences equivalent to Haar random two-qubit gates, each gate subsequence including at least one SQiSW gate;
generating a recovery gate based on the gate sequence; and
providing commands applying the gate sequence and the recovery gate to a quantum component and obtaining an output from the quantum component.
17. The method of claim 16 , wherein:
generating one of the gate subsequences comprises generating a Haar random two-qubit gate and decomposing, using two Cartan decompositions, the Haar random two-qubit gate into the one of the gate subsequences.
18. The method of claim 16 , wherein:
generating one of the gate subsequences comprises:
sampling a point in a Weyl chamber; and
generating the one of the gate subsequences based on the sampled point.
19. The method of claim 16 , wherein:
generating one of the gate subsequences comprises:
generating three interaction coefficients;
determining interleaving single qubit gates based on the three interaction coefficients;
sampling four Haar-random single-qubit gates; and
generating the one of the gate subsequences using, in part, the interleaving single qubit gates and the four sampled Haar-random single-qubit gates.
20. The method of claim 16 , wherein:
generating one of the gate subsequences comprises:
generating three interaction coefficients;
generating three canonicalized interaction coefficients and two first single-qubit gates using the three interaction coefficients;
generating two second single-qubit gates by performing a Cartan decomposition on a gate corresponding to the three canonicalized interaction coefficients;
determining second interleaving single-qubit gates based on the three canonicalized interaction coefficients; and
generating the one of the gate subsequences using, in part, the second interleaving single-qubit gates, the two first single-qubit gates, and the two second single-qubit gates.