GTM-MS3FJPH
Skip to main content

Cooperations

Recent examples (in descending order by date of operation)

Singapore Agency for Science, Technology and Research

Extension [PRIMERGY BX900]

Date operation commenced May 2011
Total theoretical peak performance 45 TFLOPS
Fields of use Fluid dynamics, materials science, etc
Purpose Development of leading edge science and technology
Features
  • Example of first overseas supercomputer system based on "PRIMERGY BX900" blade servers
  • LINPACK execution efficiency of 91.1%, the fastest in South-East Asia
  • "PRIMERGY BX900" x 450 nodes

The Research Center for Advanced Science and Technology (RCAST) at the University of Tokyo

Date operation commenced August 2010
Total theoretical peak performance 38.3 TFLOPS
Fields of use Development of drugs to treat cancer relapse and metastasis
Purpose Performing simulations to design genome-based antibody drugs that treat the relapse and metastasis of cancer
Features
  • 300 PRIMERGY BX922 S2 blade server nodes
  • Intel Xeon X5650 processors: allowing for a total of 600 CPUs
  • Enables data to be transferred at 4GB/s, while allowing for high-speed parallel calculations

Japan Atomic Energy Agency

Date operation commenced March 2010
Total theoretical peak performance 200 teraflops (note 1) (hereinafter TFLOPS)
Fields of use Research and development in the field of nuclear energy, including atomic fusion simulations
Purpose Ensure safe nuclear powered energy
Features
  • LINPACK execution performance of 186.1 TFLOPS
  • Calculation server system with three different applications
  • "PRIMERGY BX900" 2,134 nodes (4,286 CPUs, 17,072 cores)
  • "FX1" 300 nodes
  • "SPARC Enterprise M9000" 1 node

Singapore Agency for Science, Technology and Research

Date operation commenced January 2010
Total theoretical peak performance 35 TFLOPS
Fields of use Fluid dynamics, materials science, etc
Purpose Development of leading edge science and technology
Features
  • Example of first overseas supercomputer system based on "PRIMERGY BX900" blade servers
  • LINPACK execution efficiency of 91.8%, the fastest in South-East Asia
  • "PRIMERGY BX900" x 378 nodes

The Institute of Statistical Mathematics

Date operation commenced January 2010
Total theoretical peak performance 37.7 TFLOPS
Fields of use Research and development of new statistical mathematical methods such as statistical modelling, large-scale data analysis, data assimilation (next-generation simulations)
Purpose Scientific development focusing on data utilizing statistical mathematics
Features
  • Distributed memory supercomputer system (theoretical peak performance: 33.7 TFLOPS)
  • Shared memory supercomputer system (theoretical peak performance: 4 TFLOPS)
  • 35 x "ETERNUS DX80" large-scale shared storage systems
  • Highly operable implementation using the "Parallelnavi" Series of HPC middleware

RIKEN

Date operation commenced August 2009
Total theoretical peak performance 108 TFLOPS
Fields of use Wide range of fields including physics, engineering, chemistry, neuroscience, and life sciences
Purpose Science and technology research and development
Features
  • LINPACK execution performance of 97.94 TFLOPS on PC cluster system
  • Calculation server system with three different applications (super-parallel PC cluster, large capacity memory computers, multipurpose PC cluster)
  • "PRIMERGY RX200 S5", 1,024 units (2,048 CPUs, 8,192 cores)

Japan Aerospace Exploration Agency

Date operation commenced April 2009
Total theoretical peak performance 135 TFLOPS
Fields of use Research and development in the field of aerospace
Purpose Fluid simulations, etc. Mainly focused on aviation
Features
  • LINPACK execution performance of 110.6 TFLOPS with 91.19% execution efficiency
  • "FX1" 3,392 nodes

Related Products

Related Links

Footnotes

(note1) TFLOPS :

Tera Floating point number Operations Per Second