GRAPE-DR project aims to construct an information system infrastructure which combines the high-speed Internet, high-speed computer systems, and high-capacity storage systems to serve the need in the field of scientific researches. Specifically (1) distributed data sharing for experimental science, (2) high-speed computation for various large-scale numerical simulations, and (3) distributed database processing for scientific data.
By 2008, we construct an information system infrastructure for science with 2PFLOPS computation power and network throughput over 40Gbps. We also provide libraries, applications and database system with mining facilities for various fields of academic research.
(1) Distributed data sharing for experimental science, (2) high-speed computation for various large-scale numerical simulations, and (3) distributed database processing for scientific data. We employ the Data Reservoir technology with parallel IP storage for network connection. High-speed computation and data processing for numerical computation, and search and graph-matching capability, implemented through reconstruction of the high-speed simulation technology cultivated by GRAPE-6 in a super-parallel chip processor architecture with pipeline and shared memory.
The GRAPE-DR system consists of GRAPE-DR clusters located in a distant places connected by high-speed Internet with IPv4/IPv6 dual stacks with bandwidth of 40Gbps to 400Gbps level. A GRAPE-DR cluster consists of 32 to 128 PCs with GRAPE-DR card with 8 GRAPE-DR chips per each card. One cluster system has 50 to 400 TByte storage, GRAPE-DR compiler, and other system software such as operating system, WEB interface and distributed shared data system.
Applications include (1) Astrophysical particle simulation, (2) Fluid/Structural analysis by Freemesh method, (3) Molecular dynamics, (4) Protein structure analysis and similar substructure search, (5) SNPs identification by sequence matching, and (6) Genome analysis with higher-order structure search will be implemented on the above mentioned system infrastructure.
This group will construct distributed data sharing system for scientific research and develop highly efficient network communication method. National Institute of Information and Communications Technology takes charge of the network control system for data transmission and an autonomous bandwidth control system, and NTT Communications takes charge of experiments using real long line connection.
This group develops GRAPE-DR super parallel chip processor for computation, GRAPE-DR card with eight GRAPE-DR chips, and GRAPE-DR cluster system with system software such as operating system, compiler, and libraries.
This group develops softwares for astrophysical simulation, astronomical observational database, molecular dynamics simulation, genome sequence analysis, boundary element structure analysis, mass spectroscopy data search databases, and gene search problems. The National Astronomical Observatory of Japan, RIKEN and the University of Tokyo involove in this software development.