US 12,105,769 B2
Optimization problem solving calculation apparatus
Kosuke Tatsumura, Yokohama (JP); and Hayato Goto, Kawasaki (JP)
Assigned to KABUSHIKI KAISHA TOSHIBA, Tokyo (JP)
Filed by KABUSHIKI KAISHA TOSHIBA, Tokyo (JP)
Filed on Jun. 15, 2023, as Appl. No. 18/335,274.
Application 18/335,274 is a continuation of application No. 17/363,872, filed on Jun. 30, 2021, granted, now 11,720,645.
Application 17/363,872 is a continuation of application No. 16/289,084, filed on Feb. 28, 2019, granted, now 11,093,581, issued on Aug. 17, 2021.
Claims priority of application No. 2018-174270 (JP), filed on Sep. 18, 2018.
Prior Publication US 2023/0325467 A1, Oct. 12, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/16 (2006.01); G06F 17/11 (2006.01); G06F 17/13 (2006.01)
CPC G06F 17/16 (2013.01) [G06F 17/11 (2013.01); G06F 17/13 (2013.01)] 31 Claims
OG exemplary drawing
 
1. A calculation apparatus configured to solve a combinatorial optimization problem, the apparatus comprising:
input circuitry configured to acquire, as the combinatorial optimization problem, a coefficient matrix including coefficients in N rows and N columns;
matrix multiplication circuitry including a plurality of series connected block matrix multiplication circuits each including a matrix memory and a buffer outputting to a successive block matrix multiplication circuit, configured to calculate N second intermediate variables at a first time point by performing matrix multiplication between N (N is an integer equal to or greater than two) first intermediate variables at the first time point and the coefficient matrix;
time evolution circuitry configured to calculate, based on the N second intermediate variables at the first time point, N first variables at a second time point and N first intermediate variables at the second time point, the second time point being a time point following one sampling period after the first time point;
management circuitry configured to increment a time point from a start time point to an end time point by adding a sampling period for each time point and to control the matrix multiplication circuitry and the time evolution circuitry such that at each time point, the matrix multiplication circuitry calculates N second intermediate variables at the time point, and the time evolution circuitry calculates N first variables and N first intermediate variables at a subsequent incremented time point; and
output circuitry configured to output, as a solution to the combinatorial optimization problem, N first variables at a preset end time.