Supported by the National Science Foundation of China(Nos.61272097,71203064,71103077);the Natural Science Foundation of Shanghai(No.12ZR1443000);the Funding Research and Innovation Project of Shanghai Municipal Education Commission(No.12ZZ182);the Fundamental Research Funds for the Central Universitiesand the Local Colleges and Universities "1025" Connotation Construction Project of Shanghai(No.nhky-2012-10);the Foundation of Shanghai University of Engineering Science(No.A-0501-13-012)
This paper proposes a new Graphics Processing Unit(GPU)-accelerated storage format to speed up Sparse Matrix Vector Products(SMVPs) for Finite Element Method(FEM) analysis of electromagnetic problems.A new format call...