A high performance linear equation solver on the VPP500 parallel supercomputer

Conference · Fri Dec 30 23:00:00 EST 1994

OSTI ID:87681

Nakanishi, Makoto; Ina, Hiroshi ^[1]; Miura, Kenichi ^[2]

---

1. Fujitsu Ltd., Numazu (Japan). Middleware Division III
2. Fujitsu America, Inc., San Jose, CA (United States). SuperComputer Group

This paper describes the implementation of two high performance linear equation solvers developed for the Fujitsu VPP500, a distributed memory parallel supercomputer system. The solvers take advantage of the key architectural features of VPP500--(1) scalability for an arbitrary number of processors up to 222 processors, (2) flexible data transfer among processors provided by a crossbar interconnection network, (3) vector processing capability on each processor, and (4) overlapped computation and transfer. The general linear equation solver based on the blocked LU decomposition method achieves 120.0 GFLOPS performance with 100 processors in the LIN-PACK Highly Parallel Computing benchmark.

• OSTI ID: 87681
• Report Number(s): CONF-941118--; ISBN 0-8186-6605-6
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
ALGORITHMS
COMPUTER ARCHITECTURE
DATA TRANSMISSION
DISTRIBUTED DATA PROCESSING
EQUATIONS
FUJITSU COMPUTERS
MEMORY MANAGEMENT
PARALLEL PROCESSING
PERFORMANCE
SUPERCOMPUTERS