Abstract


This report documents our recent progress in developing parallel processing method for rigid body simulation using PC cluster. The parallel processing method developed in last year has some problems that the performance of the PC cluster can¡Çt be fully demonstrated and in the case that the calculation load isn¡Çt so heavy, there arise PC cluster¡Çs nodes to which no task is assigned. In this study, the parallel processing method has been improved and new function has been developed to predict future tasks and preprocess them in advance during the calculation load isn¡Çt so heavy.

Powered by JDBS Ver.2.1.0
(C) 2020 Hirotake Ishii