US 11,842,220 B2
Parallelization method and apparatus with processing of neural network model for manycore system
Jaeyeon Kim, Suwon-si (KR)
Assigned to Samsung Electronics Co., Ltd., Suwon-si (KR)
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Filed on Apr. 7, 2021, as Appl. No. 17/224,428.
Claims priority of application No. 10-2020-0141365 (KR), filed on Oct. 28, 2020.
Prior Publication US 2022/0129325 A1, Apr. 28, 2022
Int. Cl. G06F 9/50 (2006.01); G06N 3/063 (2023.01); G06F 15/80 (2006.01)
CPC G06F 9/5061 (2013.01) [G06F 15/80 (2013.01); G06N 3/063 (2013.01)] 21 Claims
OG exemplary drawing
 
1. A parallelization method comprising:
generating a profiling result by performing profiling on a target neural network based on model information of the target neural network and architecture information of a manycore system;
determining an assignment strategy to assign a plurality of cores of each of a plurality of clusters of the manycore system to a plurality of layers of the target neural network, based on the profiling result; and
generating a parallelization strategy for parallel processing of the manycore system based on the assignment strategy,
wherein the profiling result comprises a communication cost and a unit execution time,
wherein the communication cost is a cost to transmit processing results between cores of the manycore system, and
wherein the execution time is based on at least one of
a time for a single core of the manycore system to execute a single layer of the target neural network, or
a time for a single cluster of the manycore system to execute a single layer of the target neural network.