US 12,443,509 B2
Method for measuring performance of neural processing device and device for measuring performance
Jinseok Kim, Seongnam-si (KR)
Assigned to Rebellions Inc., Seongnam-si (KR)
Filed by Rebellions Inc., Seongnam-si (KR)
Filed on Jan. 10, 2024, as Appl. No. 18/409,709.
Application 18/409,709 is a continuation of application No. 18/191,731, filed on Mar. 28, 2023, granted, now 11,907,098.
Claims priority of application No. 10-2022-0041117 (KR), filed on Apr. 1, 2022; and application No. 10-2023-0025122 (KR), filed on Feb. 24, 2023.
Prior Publication US 2024/0160555 A1, May 16, 2024
Int. Cl. G06F 11/34 (2006.01); G06F 15/78 (2006.01)
CPC G06F 11/3495 (2013.01) [G06F 11/3404 (2013.01); G06F 11/3409 (2013.01); G06F 11/3447 (2013.01); G06F 15/781 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method for measuring performance of a processing device, comprising:
receiving, at a processor, hardware information associated with the processing device and a calculation task;
generating, via the processor, hardware model as agents, according to the hardware information;
generating, via the processor, an event model by dividing the calculation task by events for the agents based on the hardware model, wherein the event model includes nodes corresponding to the agents and edges corresponding to the events; and
measuring, via the processor, a total duration of the calculation task through simulation of the event model based on durations of the edges, each of the edges is data transmission between nodes, wherein the event model is regenerated after removing an edge,
wherein measuring the total duration comprises:
advancing time by a final duration of an edge that is completed earliest among the edges;
removing the edge that is completed the earliest;
computing a final duration of each of all the edges again; and
measuring the total duration by summing the advanced time if all the edges have been removed.