Lawrence Berkeley National Laboratory

The last decade has witnessed a rapid proliferation of superscalar cache-based microprocessors to build high-endcomputing (HEC) platforms, primarily because of their generality,scalability, and cost effectiveness. However, the growing gap between sustained and peak performance for full-scale scientific applications on such platforms has become major concern in high performance computing. The latest generation of custom-built parallel vector systems have the potential to address this concern for numerical algorithms with sufficient regularity in their computational structure. In this work, we explore two and three dimensional implementations of a plasma physics application, as well as a leading astrophysics package on some of today's most powerful supercomputing platforms. Results compare performance between the the vector-based Cray X1, EarthSimulator, and newly-released NEC SX- 8, with the commodity-based superscalar platforms of the IBM Power3, Intel Itanium2, and AMDOpteron. Overall results show that the SX-8 attains unprecedented aggregate performance across our evaluated applications.