Granted Patent B1

US 11,436,519 · App. 17/560,816 · Granted Sep 6, 2022

Inventors: Raouf Dridi (Leesburg, VA); Uchenna Chukwu (Leesburg, VA); Jesse Berwald (Leesburg, VA)

Assignee: Quantum Computing Inc.

G06N10/60G06N20/00

Abstract

Claims (53)

1. A method, comprising:

obtaining, with a computer system, a first set of raw outputs from a first plurality of shots run by a quantum processor processing a first part of a problem;

obtaining, with the computer system, a second set of raw outputs from a second plurality of shots run by the quantum processor or another quantum processor processing a second part of the problem;

statistically aggregating, with the computer system, the first set of raw outputs and the second set of raw outputs to determine an expectation value solving or approximating a solution to the problem, wherein:

statistically aggregating comprises determining the expectation value based on a probability distribution of at least some of the raw outputs, and

determining the expectation value comprises determining a subset of the raw outputs having greater than a threshold frequency of occurrence in the raw outputs and determining a measure of central tendency based on the subset; and

storing, with the computer system, the expectation value in memory and determining, with the computer system, based on the expectation value, a direction in parameter space in which to adjust parameters to further optimize an objective function corresponding to the problem, the parameters being classical parameters to be applied to quantum circuits of the quantum processor or the another quantum processor.

2. The method of claim 1 , wherein determining the direction in parameter space in which to adjust parameters comprises determining the direction in parameter space in which to adjust parameters to locally further optimize the objective function.

3. The method of claim 1 , wherein:

the first plurality of shots is greater than 100;

the first set of raw outputs comprises more than 5 values per output; and

each shot in the first plurality of shots is a quantum circuit repetition.

4. The method of claim 1 , wherein statistically aggregating comprises:

grouping raw outputs in the first set of raw outputs that are identical with one another and determining respective counts of members for each group; and

grouping raw outputs in the second set of raw outputs that are identical with one another and determining respective counts of members for each group.

5. The method of claim 1 , wherein determining the expectation value based on a probability distribution of raw outputs comprises:

concatenating raw outputs in the first set of raw outputs with raw outputs in the second set of raw outputs by shot number, such that raw outputs having the same shot number are concatenated with one another to form concatenated raw outputs; and

determining the expectation value based on a probability distribution of the concatenated raw outputs.

6. The method of claim 5 , wherein determining the measure of central tendency comprises determining a measure of central tendency of at least some of the concatenated raw outputs.

7. The method of claim 5 , wherein determining a subset of the raw outputs having greater than a threshold frequency of occurrence in the raw outputs comprises determining a subset of the concatenated raw outputs having greater than a threshold frequency of occurrence in the concatenated raw outputs and determining a measure of central tendency based on the subset of the concatenated raw outputs.

8. The method of claim 1 , wherein statistically aggregating comprises steps for statistically aggregating raw outputs.

9. The method of claim 1 , wherein determining the direction in parameter space in which to adjust parameters comprises empirically estimating partial derivatives of the parameters with respect to the expectation value with parameter shifts.

10. The method of claim 1 , wherein determining the direction in parameter space in which to adjust parameters comprises empirically estimating partial derivatives of the parameters with respect to the expectation value with ancilla circuits.

11. The method of claim 1 , wherein processing a first part of the problem comprises processing a first set of inputs and processing a second part of the problem comprises processing a second set of inputs, wherein the first set of inputs, the second set of inputs, or a combination thereof comprise the parameters.

12. The method of claim 1 , wherein determining a direction in parameter space in which to adjust the parameters further comprises adjusting, with the computer system, at least one parameter based on the expectation value.

13. The method of claim 12 , wherein storing the expectation value in memory further comprises storing the adjustment in memory.

14. The method of claim 12 , further comprising:

obtaining, with the computer system, a third set of raw outputs from a third plurality of shots run by a quantum processor processing a first adjusted part of the problem;

obtaining, with the computer system, a fourth set of raw outputs from a fourth plurality of shots run by the quantum processor or another quantum processor processing a second adjusted part of the problem;

statistically aggregating, with the computer system, the third set of raw outputs and the fourth set of raw outputs to determine an second expectation value solving or approximating a solution to the problem;

adjusting, with the computer system, at least one parameter based on the second expectation value; and

storing, with the computer system, the second expectation value in memory.

15. The method of claim 1 , comprising steps for determining adjustments to parameters associated with qubits.

16. A non-transitory, computer-readable medium storing instructions that when executed effectuate operations comprising:

obtaining, with a computer system, a first set of raw outputs from a first plurality of shots run by a quantum processor processing a first part of a problem;

obtaining, with the computer system, a second set of raw outputs from a second plurality of shots run by the quantum processor or another quantum processor processing a second part of the problem;

statistically aggregating, with the computer system, the first set of raw outputs and the second set of raw outputs to determine an expectation value solving or approximating a solution to the problem, wherein:

statistically aggregating comprises determining the expectation value based on a probability distribution of the raw outputs, and

determining the expectation value comprises determining a subset of the raw outputs having greater than a threshold frequency of occurrence in the raw outputs and determining a measure of central tendency based on the subset; and

storing, with the computer system, the expectation value in memory and determining, with the computer system, based on the expectation value, a direction in parameter space in which to adjust parameters to further optimize an objective function corresponding to the problem, the parameters being classical parameters to be applied to quantum circuits of the quantum processor or the another quantum processor.

17. A method comprising:

obtaining, with a computer system, a first set of raw outputs from a first plurality of shots run by a quantum processor processing a first part of a problem;

obtaining, with the computer system, a second set of raw outputs from a second plurality of shots run by the quantum processor or another quantum processor processing a second part of the problem;

statistically aggregating, with the computer system, the first set of raw outputs and the second set of raw outputs to determine an expectation value solving or approximating a solution to the problem,

wherein statistically aggregating comprises:

ranking raw outputs in the first set of raw outputs according to frequency of occurrence in the first set of raw outputs to form a first ranking of raw outputs;

ranking raw outputs in the second set of raw outputs according to frequency of occurrence in the second set of raw outputs to form a second ranking of raw outputs; and

concatenating raw outputs in the first ranking of raw outputs with raw outputs in the second ranking of raw outputs according to rank order to form concatenated raw outputs;

determining the expectation value based on a probability distribution of the concatenated raw outputs,

wherein determining the expectation value comprises determining a subset of the raw outputs having greater than a threshold frequency of occurrence in the raw outputs and determining a measure of central tendency based on the subset; and

storing, with the computer system, the expectation value in memory and determining, with the computer system, based on the expectation value, a direction in parameter space in which to adjust parameters to further optimize an objective function corresponding to the problem, the parameters being classical parameters to be applied to quantum circuits of the quantum processor or the another quantum processor.

18. The method of claim 17 , wherein determining the expectation value based on the probability distribution of the concatenated raw outputs comprises selecting a most common concatenated raw output as the expectation value.

19. The method of claim 17 , wherein determining the expectation value based on the probability distribution of the concatenated raw outputs comprises determining a measure of central tendency from concatenated raw outputs having greater than a threshold frequency of occurrence in the concatenated raw outputs.

Assignments (3)

Cited By (10)

US 12,223,343 US 12,346,407 US 12,387,121 US 12,395,326 US 12,456,066 US 12,530,730 US 12,547,466 US 12,632,767 US 12,645,761 US 12,664,234