IP Library Granted Patent US 12,561,599
Granted Patent B2
US 12,561,599 · App. 17/959,962 · Granted Feb 24, 2026

Network depth-minimizing quantum compiler

Inventors: Thomas Fischbacher (Thalwil, CH); Luca Versari (Zurich, CH)
Assignee: Google LLC
G06N10/80
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,561,599
App. No.
17/959,962
Granted
Feb 24, 2026
Kind
B2
Abstract

Implementations disclosed describe techniques used for compiling a quantum algorithm for execution on a plurality of quantum circuits, including accessing, by a processing device, the quantum algorithm, identifying a matrix associated with the quantum algorithm, determining a representation of the identified matrix as a matrix decomposition that includes a plurality of transformation matrices, wherein one or more of the plurality of transformation matrices perform multiple instances of two-dimensional rotations; and generating a circuit map that maps execution of the matrix decomposition on the plurality of quantum circuits.

Claims (44)

1 . A method of compiling a quantum algorithm for execution on a plurality of quantum circuits, the method comprising:

accessing, by a processing device, the quantum algorithm;

identifying, by the processing device, a matrix associated with the quantum algorithm;

determining, by the processing device, a representation of the identified matrix as a matrix decomposition comprising a matrix product of the matrix and a plurality of transformation matrices (TMs), wherein each TM of at least a subset of one or more of the plurality of TMs performs multiple instances of complex two-dimensional (2D) rotations that reduce to zero multiple matrix elements of a common column of the matrix product;

generating, by the processing device, a circuit map that maps execution of the matrix decomposition on the plurality of quantum circuits; and

executing, using the circuit map, the quantum algorithm on the plurality of quantum circuits.

2 . The method of claim 1 , wherein the matrix associated with the quantum algorithm is a unitary matrix.

3 . The method of claim 1 , wherein the matrix associated with the quantum algorithm has a dimension N×N and wherein the plurality of TMs comprises a TM that performs one of: i) N/2 of the instances of 2D rotations, wherein Nis even, or ii) (N−1)/2 of the instances of 2D rotations, wherein Nis odd.

4 . The method of claim 1 , wherein the matrix associated with the quantum algorithm has a dimension N×N, wherein a subset of the plurality of TMs eliminates all non-diagonal elements of a first column or a last column of the matrix associated with the quantum algorithm, and wherein a first subset of the plurality of TMs has a number of TMs that is at most one more than a binary logarithm of N.

5 . The method of claim 1 , wherein each of the multiple instances of the 2D rotations of a first TM of the plurality of TMs is associated with a respective pair of rows of the first TM, and wherein the respective pair of rows is selected based on a magnitude of each of matrix elements of a predetermined partition of the matrix associated with the quantum algorithm.

6 . The method of claim 5 , wherein the predetermined partition is one of a first column, a last column, a first row, or a last row of i) the matrix associated with the quantum algorithm or ii) a rearranged matrix obtained by rearranging elements of the matrix associated with the quantum algorithm.

7 . The method of claim 1 , wherein the circuit map comprises an interconnected arrangement of the plurality of quantum circuits, each of the plurality of quantum circuits capable of performing one or more 2D rotations.

8 . The method of claim 1 , wherein the circuit map comprises one or more parameters for each of the plurality of quantum circuits.

9 . The method of claim 8 , wherein a first quantum circuit of the plurality of quantum circuits comprises an optical circuit having a first input port, a second input port, a first output port, and a second output port, and wherein the one or more parameters for the first quantum circuit determine at least:

an amplitude for an optical beam, entering the optical circuit through the first input port, to exit the optical circuit through the first output port; and

a phase shift acquired by the optical beam entering the optical circuit through the first input port and exiting the optical circuit through one of the first output port or the second output port.

10 . A system comprising:

a memory; and

a processing device, communicatively coupled to the memory, configured to:

access a quantum algorithm;

identify a matrix associated with the quantum algorithm;

determine a representation of the identified matrix as a matrix decomposition comprising a matrix product of the matrix and a plurality of transformation matrices (TMs), wherein each T M of at least a subset of one or more of the plurality of TMs performs multiple instances of complex two-dimensional (2D) rotations that reduce to zero multiple matrix elements of a common column of the matrix product;

generate a circuit map that maps execution of the matrix decomposition on a plurality of quantum circuits; and

execute, using the circuit map, the quantum algorithm on the plurality of quantum circuits.

11 . The system of claim 10 , wherein the matrix associated with the quantum algorithm has a dimension N×N and wherein the plurality of TMs comprises a TM that performs one of: i) N/2 of the instances of 2D rotations, wherein Nis even, or ii) (N−1)/2 of the instances of 2D rotations, wherein Nis odd.

12 . The system of claim 10 , wherein the matrix associated with the quantum algorithm has a dimension N×N, wherein a subset of the plurality of TMs eliminates all non-diagonal elements of a first column or a last column of the matrix associated with the quantum algorithm, and wherein the subset of the plurality of TMs has a number of TMs that is at most one more than a binary logarithm of N.

13 . The system of claim 10 , wherein each of the multiple instances of the 2D rotations of a first TM of the plurality of TMs is associated with a respective pair of rows of the first TM, and wherein the respective pair of rows is selected based on a magnitude of each of matrix elements of a predetermined partition of the matrix associated with the quantum algorithm.

14 . The system of claim 13 , wherein the predetermined partition of the first TM is one of a first column, a last column, a first row, or a last row of the matrix associated with the quantum algorithm.

15 . The system of claim 10 , wherein the circuit map comprises an interconnected arrangement of the plurality of quantum circuits, each of the plurality of quantum circuits capable of performing one or more 2D rotations.

16 . The system of claim 10 , wherein the circuit map comprises one or more parameters for each of the plurality of quantum circuits.

17 . The system of claim 16 , wherein a first quantum circuit of the plurality of quantum circuits comprises an optical circuit having a first input port, a second input port, a first output port, and a second output port, and wherein the one or more parameters for the first quantum circuit determine at least:

an amplitude for an optical beam, entering the optical circuit through the first input port, to exit the optical circuit through the first output port; and

a phase shift acquired by the optical beam entering the optical circuit through the first input port and exiting the optical circuit through one of the first output port or the second output port.

18 . The system of claim 10 , wherein to identify the matrix associated with the quantum algorithm, the processing device is configured to access a plurality of elements of the matrix stored in the memory.

19 . A non-transitory computer-readable medium to store instructions, which when executed by a processing device, cause the processing device to:

access a quantum algorithm;

identify a matrix associated with the quantum algorithm;

determine a representation of the identified matrix as a matrix decomposition comprising a matrix product of the matrix and a plurality of transformation matrices (TMs), wherein each TM of at least a subset of one or more of the plurality of TMs performs multiple instances of complex two-dimensional (2D) rotations that reduce to zero multiple matrix elements of a common column of the matrix product;

generate a circuit map that maps execution of the matrix decomposition on a plurality of quantum circuits; and

execute, using the circuit map, the quantum algorithm on the plurality of quantum circuits.

20 . The non-transitory computer-readable medium of claim 19 , wherein the matrix associated with the quantum algorithm has a dimension N×N and wherein the plurality of TMs comprises a TM that performs one of: i) N/2 of the instances of 2D rotations, wherein Nis even, or ii) (N−1)/2 of the instances of 2D rotations, wherein Nis odd.

21 . The non-transitory computer-readable medium of claim 19 , wherein the circuit map comprises:

an interconnected arrangement of the plurality of quantum circuits, each of the plurality of quantum circuits capable of performing one or more 2D rotations, and

one or more parameters for each of the plurality of quantum circuits.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 7, 2022
From: FISCHBACHER, THOMAS; VERSARI, LUCA
To: GOOGLE LLC
Reel/Frame 061351/0187 →
Continuity (2)
Provisional Application 63262320 · Oct 8, 2021
Related Publication 20230115065A1 · Apr 13, 2023
References Cited (13)
US 9791258B2 · Mower · 2017 [cited by examiner]
US 11386347B2 · Low · 2022 [cited by examiner]
US 12020117B2 · Javadiabhari · 2024 [cited by examiner]
US 20060123363A1 · Williams · 2006 [cited by examiner]
US 20210096443A1 · Dhand · 2021 [cited by examiner]
US 20230115065A1 · Fischbacher · 2023 [cited by examiner]
Mottonen, “Decompositions of general quantum gates”, (hereinafter Mottonen), Apr. 14, 2005, pp. 1-25, [online][retrieved on Jun. 25, 2025]. Retrieved from < https://arxiv.org/pdf/quant-ph/0504100> (Year: 2005). [cited by examiner]
N. Tischler, “Quantum Optical Realization of Arbitrary Linear Transformations Allowing for Loss and Gain”, Apr. 13, 2018, 13 pages, [online], [retrieved on Jun. 25, 2025]. Retrieved from <https://journals.aps.org/prx/ab… [cited by examiner]
PCT International Search Report and Written Opinion for International Application No. PCT/US2022/045939, mailed Jan. 27, 2023, 16 pages. [cited by applicant]
Nora Tischler et al: “Quantum optical realization of arbitrary linear transformations allowing for loss and gain”, arxiv.org, Cornell University Library, 201 Olin Library Cornell University Ithaca, NY 14853, Apr. 15, 20… [cited by applicant]
Fischbacher et al., “Single-Photon Image Classification,” Published as a conference paper at International Conference on Learning Representations (ICLR), Mar. 12, 2021, 14 pages. [cited by applicant]
Krol et al,. “Efficient decomposition of unitary matrices in quantum circuit compilers,” arXiv:2101.02993v1, Jan. 8, 2021, 13 pages. [cited by applicant]
Reck et al., “Experimental Realization of Any Discrete Unitary Operator,” The American Physical Society, vol. 73, No. 1, Jul. 4, 1994, 6 pages. [cited by applicant]