Systems and methods for degeneracy mitigation in a quantum processor
Degeneracy in analog processor (e.g., quantum processor) operation is mitigated via use of floppy qubits or domains of floppy qubits (i.e., qubit(s) for which the state can be flipped with no change in energy), which can significantly boost hardware performance on certain problems, as well as improve hardware performance for more general problem sets. Samples are drawn from an analog processor, and devices comprising the analog processor evaluated for floppiness. A normalized floppiness metric is calculated, and an offset added to advance the device in annealing. Degeneracy in a hybrid computing system that comprises a quantum processor is mitigated by determining a magnetic susceptibility of a qubit, and tuning a tunneling rate for the qubit based on a tunneling rate offset determined based on the magnetic susceptibility. Quantum annealing evolution is controlled by causing the evolution to pause for a determined pause duration.
1. A method of selecting an annealing schedule for a problem in a hybrid computing system, the hybrid computing system comprising a quantum processor and a digital processor, the quantum processor and the digital processor communicatively coupled to one another, the quantum processor comprising a plurality of qubits, the method comprising:
generating a set of candidate annealing schedules comprising a plurality of candidate annealing schedules by the digital processor;
selecting an annealing schedule from the set of candidate annealing schedules by the digital processor based on an objective function which receives each annealing schedule from the set of candidate annealing schedules as input and provides a measure of at least one characteristic of each input annealing schedule as output to be used in a selection criteria;
sending the problem to the quantum processor by the digital processor; and
executing the problem at the quantum processor according to the selected annealing schedule.
2. The method of claim 1 wherein the method comprises selecting the objective function from a set of one or more objective functions.
3. The method of claim 1 wherein generating the set of annealing schedules comprises performing an optimization algorithm based on the objective function.
4. The method of claim 3 wherein performing the optimization algorithm comprises performing parallel tempering.
5. The method of claim 4 wherein performing parallel tempering comprises measuring, with the objective function, at least one of: models and chains linking models generated by applying parallel tempering to the problem as modified by the input annealing schedule.
6. The method of claim 5 wherein measuring with the objective function comprises computing at most a threshold number of parallel tempering iterations, the threshold number less than a number of iterations for solving the problem.
7. The method of claim 3 wherein performing the optimization algorithm comprises performing Bayesian optimization.
8. The method of claim 7 wherein performing Bayesian optimization comprises measuring, with the objective function, a ground state distribution of the problem according to the input annealing schedule.
9. The method of claim 8 wherein measuring with the objective function comprises computing at least one of: a measure of similarity between a ground state distribution and a characteristic of one or more outliers of the ground state distribution.
10. The method of claim 9 wherein measuring with the objective function comprises measuring an entropy of the ground state distribution.
11. The method of claim 9 wherein measuring with the objective function comprises measuring a distance of the ground state distribution from a uniform distribution according to a distance metric.
12. The method of claim 9 wherein measuring with the objective function comprises measuring a Gini coefficient of the ground state distribution.
13. The method of claim 9 wherein measuring with the objective function comprises measuring a ratio of a maximum to a minimum probability of the ground state distribution.
14. The method of claim 1 wherein selecting the annealing schedule from the one or more annealing schedules comprises determining that the annealing schedule provides an optimal result relative to the one or more annealing schedules.
15. The method of claim 14 wherein generating the set of annealing schedules comprises generating a plurality of annealing schedules, selecting an interim annealing schedule based on the objective function, and generating the set of annealing schedules based on the interim annealing schedule.
16. A hybrid computing system comprising a quantum processor and a digital processor, the quantum processor and the digital processor communicatively coupled to one another, the quantum processor comprising a plurality of qubits, the hybrid computing system operable to:
generate a set of candidate annealing schedules for a problem, the set of candidate annealing schedules comprising a plurality of candidate annealing schedules;
select an annealing schedule from the set of candidate annealing schedules based on an objective function which receives each annealing schedule from the set of candidate annealing schedules as input and provides a measure of at least one characteristic of each input annealing schedule as output to be used in a selection criteria;
send the problem to the quantum processor by the digital processor; and
execute the problem at the quantum processor according to the selected annealing schedule.
17. The hybrid computing system of claim 16 wherein the hybrid computer generates the set of candidate annealing schedules with an optimization algorithm based on the objective function.
18. The hybrid computing system of claim 17 wherein the optimization algorithm comprises one of parallel tempering and Bayesian optimization.
19. The hybrid computing system of claim 16 wherein the selected annealing schedule from the set of annealing schedules provides an optimal result relative to the set of annealing schedules.
20. The hybrid computing system of claim 16 wherein the hybrid computer generates the set of annealing schedules based on an interim annealing schedule, wherein the hybrid computer generates a plurality of annealing schedules, and the hybrid computing system is operable to select an interim annealing schedule from the plurality of annealing schedules based on the objective function, and generates the set of annealing schedules based on the interim annealing schedule.