Parallel Performance of Tectonic Loading Process Model at Transcurrent Plate Boundaries

ABSTRACT Code for simulation of tectonic loading process at transcurrent plate boundaries was parallelized by MPI. Matrix assembly and inversion processes have been largely optimized for parallel computing, and 2-way parallelization was developed for local operations for parameter points and data/integral points. Computations using this method on a Hitachi SR2201 with up to 128 processors demonstrated good scalability. REPORT DATE October 30th, 2001. TOTAL NO. OF PAGES 8 ANY OTHER IDENTIFYING INFORMATION OF THIS REPORT Original paper presented at "The 2nd ACES (APEC Cooperation for Earthquake Simulation) Workshop", October 15-20, 2000, Tokyo and Hakone, Japan. Additional computations by Hitachi SR2201 and paper was revised in September, 2001. Research Organization for Information Science & Technology (RIST) 2-2-54, Nakameguro, Meguro-ku Tokyo, 153-0061, Japan TEL: +81-3-3712-5321 FAX: +81-3-3712-5552 http://geofem.tokyo.rist.or.jp/ Copyright©2001 Research Organization for Information Science & Technology (RIST)