Abstract: High-performance cluster computing systems (HPCC) is widely accepted because of its powerful computing capability, high I/O performance, and strong performance management system scalability. In this paper, the development and topological composition of high performance computing system are briefly introduced. System environment, performance features, HPC management and its applications of this supercomputational platform are comparably described in detail. As an example, we presented the system constructed in IES consisted of 192 computational nodes, 4 I/O nodes, 1 login node, 1 management node, 1 windows cluster service node, a group of disk array, a group of computing, I/O management and network. The China TOP100 Rank of High Performance Computer of IES released by the Specialty Association of Mathematical & Scientific Software (SAMSS),CSIA is listed in 36 and 63 respectively in 2009 and 2010, which proved that it has competitive capacity and can provide powerful support for earthquake scientific research tasks in recent years.

Key words: Topological Architecture, System Environmant, Cluster Management, High Performance Computing System