Study on GPU-accelerated extraction of interconnects parasitic using CUDA and MPI
Institute of Electr. Eng., Chinese Academy of Science, Beijing, China
14th Biennial IEEE Conference on Electromagnetic Field Computation (CEFC), 2010
Parallel computation is application-oriented, particularly for the GPU (Graphics Processing Unit) with the inherent parallelism. This paper shows the architecture of a GPU cluster based on MPI (Message Passing Interface) and CUDA (Compute Unified Device Architecture). Results show that the acceleration ratio is obviously improved but the acceleration effect seems decelerated in large-scale GPU cluster. The parallel algorithm is mainly focused on task partitioning sparse matrix-vector multiplications (SpVM) in GPUs.
March 6, 2011 by hgpu