Parallel ICCG Algorithm in Finite Element Analysis Using Domain Division on a Shared Memory Multiprocessor

摘要

A parallel ICCG algorithm on a shared memory multiprocessor is presented for finite element analysis using domain division idea. It is simple and easy to program based on the original sequent algorithm and its speedup ratio is good. It has been implemented in parallel on a multiprocessor with 8 CPUs, SGI Challenge-L.