# US 2019/0317978 A1 · Re-equilibrated quantum sampling

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0317978 A1
- Document ID: US20190317978A1
- Status: Published application
- Kind: A1
- Application number: 16/433,027
- Filed: 2019-06-06
- Published: 2019-10-17
- Earliest claimed: 2017-01-12
- Assignees: D-Wave Systems Inc.
- Applicants: D-Wave Systems Inc.
- Inventors: Mohammad H. Amin; Evgeny A. Andriyash
- Web page: https://patentgrep.com/patent/US20190317978A1

## CPC

- G06F 17/17: Function evaluation by approximation methods, e.g. inter- or extrapolation, smoothing, least mean square method ({G06F17/18 takes precedence } ; interpolation for numerical control G05B19/18)
- G06F 5/01: for shifting, e.g. justifying, scaling, normalising {(digital stores in which the information is moved stepwise, e.g. shift-registers G11C19/00; digital stores in which the information circulates G11C21/00)}
- G06F 9/30145: {Instruction analysis, e.g. decoding, instruction word fields}
- G06F 17/13: Differential equations (using digital differential analysers G06F7/64)
- G06F 17/18: for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)}
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 10/40: Physical realisations or architectures of quantum processors or components for manipulating qubits, e.g. qubit coupling or qubit control
- G06N 10/60: Quantum algorithms, e.g. based on quantum optimisation, quantum Fourier or Hadamard transforms
- H04L 9/0852: {Quantum cryptography (transmission systems employing electromagnetic waves other than radio waves, e.g. light, infrared H04B10/00; wavelength-division multiplex systems H04J14/02; WDM arrangements H04J14/03)}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

The systems, devices, articles, and methods generally relate to sampling from an available probability distribution. The samples maybe used to create a desirable probability distribution, for instance for use in computing values used in computational techniques including: Importance Sampling and Markov chain Monte Carlo systems. An analog processor may operate as a sample generator, for example by: programming the analog processor with a configuration of the number of programmable parameters for the analog processor, which corresponds to a probability distribution over qubits of the analog processor, evolving the analog processor with a fast ramp operation, and reading out states for the qubits. The state for the qubits may be post processes and/or used to calculate importance weights.

## Claims

### Claim 1

A method of operation of a hybrid computing system to generate approximate equilibrium statistics, the hybrid computing system comprising an analog processor and a digital processor, the method comprising:
initializing a counter i;
determining an intermediate value si* of an annealing parameter, the annealing parameter normalized between a value of zero and a value of one;
determining a ramp time tramp wherein the ramp time tramp is smaller than a relaxation time of the analog processor and greater than a time needed for non-adiabatic transition to higher energy states;
initializing, the analog processor to a state i;
until an exit condition is met, iteratively:
evolving the analog processor from a value of one for the annealing parameter until the annealing parameter is equal to the intermediate value si*, wherein evolving the analog processor from a value of one for the annealing parameter until the annealing parameter is equal to the intermediate value si* occurs in the ramp time tramp;
holding the annealing parameter of the analog processor to the intermediate value si* for a defined dwell time;
evolving the analog processor from the intermediate value si* of the annealing parameter until the annealing parameter is equal to one, wherein evolving the analog processor from the intermediate value si* of the annealing parameter until the annealing parameter is equal to one occurs in the ramp time tramp and after the ramp time tramp the analog processor is in a state i+1; and
reading out the state i+1 of the analog processor;
if the exit condition is not met, determining a new intermediate value si+1*>si* of the annealing parameter; and
incrementing the counter i.

### Claim 2 (depends on 1)

The method of operation of claim 1 wherein determining an intermediate value si* of an annealing parameter, the annealing parameter normalized between a value of zero and a value of one comprises determining a value of the annealing parameter where the evolution of the analog processor starts deviating toward a distribution that is not a Boltzmann distribution.

### Claim 3 (depends on 1)

The method of claim 1 wherein the dwell time is a time greater than a time needed for the evolution of the analog processor to recover from any distortion caused by evolving the analog processor by the ramp time tramp.

### Claim 4 (depends on 1)

The method of claim 1 wherein reading out the state i+1 comprises reading the state i+1 with a value different than another value of the state i.

### Claim 5 (depends on 4)

The method of claim 4 wherein reading out the state i+1 comprises sampling from a distribution obtained from an evolution in the ramp time tramp from a thermal equilibrium distribution
P
β

[
z
]
=
e
-
β


E

(
z
)
Z
β
.

### Claim 6 (depends on 1)

The method of claim 1, further comprising: after reading out the state i+1 of the analog processor, post-processing the state i+1 via the digital processor to obtain a post-processed state i+1 of the analog processor.

### Claim 7 (depends on 6)

The method of claim 6 wherein post-processing the state i+1 of the analog processor comprises applying one or more passes of analog Monte Carlo simulations to the state i+1.

### Claim 8 (depends on 6)

The method of claim 6, further comprising:
assigning importance weights to the post-processed state i+1 of the analog processor via the digital processor; and
initializing, by the digital processor, the analog processor to the post-processed state i+1.

### Claim 9 (depends on 1)

The method of claim 1, further comprising:
after reading out the state i+1 of the analog processor storing the state i+1 via the digital processor; and
after meeting the exit condition, post processing the states of the analog processor via the digital processor to obtain importance weights.

### Claim 10 (depends on 9)

The method of claim 9 wherein post processing the state of the analog processor via the digital processor comprises performing one or more passes of analog Monte Carlo post processing.

### Claim 11

A hybrid computing system comprising:
a digital processor in communication with an analog processor;
at least one nontransitory processor-readable medium communicatively coupled to the digital processor, the at least one nontransitory processor-readable medium storing at least one of processor-executable instructions or data which, when executed by the digital processor, cause the digital processor to:
initialize a counter i;
determine an intermediate value si* of an annealing parameter, the annealing parameter normalized between a value of zero and a value of one;
determine a ramp time tramp wherein the ramp time tramp is smaller than a relaxation time of the analog processor and greater than a time needed for non-adiabatic transition to higher energy states;
initialize the analog processor to a state i;
until an exit condition is met, iteratively:
evolve the analog processor from a value of one for the annealing parameter until the annealing parameter is equal to the intermediate value si*, wherein causing the digital processor to evolve the analog processor from a value of one for the annealing parameter until the annealing parameter is equal to the intermediate value si* occurs in the ramp time tramp;
hold the annealing parameter of the analog processor to the intermediate value si* for a defined dwell time;
evolve the analog processor from the intermediate value si* of the annealing parameter until the annealing parameter is equal to one, wherein causing the digital processor to evolve the analog processor from the intermediate value si* of the annealing parameter until the annealing parameter is equal to one occurs in the ramp time tramp and after the ramp time tramp the analog processor is in a state i+1; and
read out the state i+1 of the analog processor;
if the exit condition is not met, determine a new intermediate value si+1*>si* of the annealing parameter; and
increment the counter i.

### Claim 12 (depends on 11)

The computing system of claim 11 wherein causing the digital processor to determine an intermediate value si* of an annealing parameter, the annealing parameter normalized between a value of zero and a value of one comprises causing the digital processor to determine a value of the annealing parameter where the evolution of the analog processor starts deviating toward a distribution that is not a Boltzmann distribution.

### Claim 13 (depends on 11)

The computing system of claim 11 wherein the dwell time is a time greater than a time needed for the evolution of the analog processor to recover from a distortion, if any, caused by evolving the analog processor by the ramp time tramp.

### Claim 14 (depends on 11)

The computing system of claim 11 wherein the state i+1 is not equal to the state i.

### Claim 15 (depends on 14)

The computing system of claim 14 wherein the state i+1 is a sample from a distribution obtained from an evolution in the ramp time tramp from a thermal equilibrium distribution
P
β

[
z
]
=
e
-
β


E

(
z
)
Z
β
.

### Claim 16 (depends on 11)

The computing system of claim 11, further comprising: after causing the digital processor to read out the state i+1 of the analog processor, causing the digital processor to post-process the state i+1 of the analog processor to obtain a post-processed state i+1 of the analog processor.

### Claim 17 (depends on 16)

The computing system of claim 16 wherein causing the digital processor to post-process the state i+1 of the analog processor comprises causing the digital processor to apply one or more passes of analog Monte Carlo simulations to the state i+1.

### Claim 18 (depends on 16)

The computing system of claim 16 wherein the at least one of processor-executable instructions or data comprise further processor-executable instructions or data which, when executed by the digital processor, cause the digital processor to:
assign importance weights to the post-processed state i+1 of the analog processor; and
initialize the analog processor to the post-processed state i+1.

### Claim 19 (depends on 11)

The computing system of claim 11 wherein the at least one of processor-executable instructions or data comprise further processor-executable instructions or data which, when executed by the digital processor, cause the digital processor to:
after reading out the state i+1 of the analog processor, store the state i+1; and
after meeting the exit condition, post process the states of the analog processor to obtain importance weights.

### Claim 20 (depends on 19)

The computing system of claim 19 wherein causing the digital processor to post process the state of the analog processor comprises causing the digital processor to perform one or more passes of analog Monte Carlo post processing.
