The PACS-CS Project aims to advance frontiers of computational sciences through development of a massively parallel cluster PACS-CS (Parallel Array Computer System for Computational Sciences) and its concentrated usage on the problems that require large-scale simulations. One of the emphases of the Project is placed on the development of first-principles quantum simulations in materials and life sciences capable to treat O(10,000) atoms, thereby allowing an exploration of the connection between the spatial structure and function characteristic of nano materials such as carbon nano-tubes and large bio molecules such as proteins. Another emphasis is advancement of full QCD simulations in particle physics and astrophysics research which forms the core of our fundamental understanding of the creation and history of our Universe.
The planning of the project was started in the early summer of 2003. The application for funding was initiated in the fall of 2003. Several hearings by the Ministry of Education, Culture, Sports, Science and Technology was conducted from January through June 2004.
The Project was formally submitted from University of Tsukuba to the Ministry of Education, Culture, Sports, Science and Technology as a 3-year proposal with a funding request to Special Fund for Education and Research for the Japanese Fiscal 2005. The Official name of the Project is “Discovery, Synthesis and Emergence of Novel Knowledge through Computational Sciences”.
The Project was formally approved by the government and was included in the Government Budget Proposal to the Diet in December 2004. The Project formally started in April 2005, and is to continue for three years through Japanese Fiscal 2007.
The Project is based at and carried out by Center for Computational Sciences (CCS), University of Tsukuba. The organization of the Project is as follows.
|Akira Ukawa (Director of CCS)|
PACS-CS System Development
|Members||Mitsuhisa Sato1),a), Daisuke Takahashi1),a)|
Materials and Life Science Research
|Members||Susumu Okada2),b), Mauro Boero2),c), Masaru Tateno2),c), Kazuhiro Yabana2),c), Takashi Nakatsukasa2),c), Junichi Iwata2), Kenji Shiraishib)|
Particle Physics Research
|Members||Tomoteru Yoshie3),c), Naruhito Ishizuka3),c), Yoshinobu Kuramashi3),c), Yusuke Taniguchi3),c), Kiyoshi Sasaki3), Tomomi Ishikawa3), Sinya Aokic), Kazuyuki Kanayac)|
|Members||Taishi Nakamoto3),c), Hiroyuki Hirashita3),c), Yoshiaki Kato3), Kanako Sugimoto3)|
Geoenvironmental and Biology Research
|Members||Yuji Inagaki4),e), Yukio Hashimotoe)|
Computational Media Research
|Members||Yoshinari Kameda5),f), Itaru Kitahara5),f)|
Computational Intelligence Research
|Members||Yoshiharu Ishikawa5),a), Toshiyuki Amagasa5),a)|
1) Division of High Performance Computing Systems, CCS
2) Division of Materials and Life Sciences, CCS
3) Division of Particle Physics and Astrophysics, CCS
4) Division of Global Environment and Biological Sciences, CCS
5) Division of Computational Informatics
a) Graduate School of Systems and Information Engineering, School of Computer Science
b) Graduate School of Pure and Applied Sciences, School of Frontier Science
c) Graduate School of Pure and Applied Sciences, School of Physics
d) Graduate School of Life and Environmental Sciences, School of Geoenvironmental Sciences
e) Graduate School of Life and Environmental Sciences, School of Structural Biosciences
f) Graduate School of Systems and Information Engineering, School of Advanced Engineering Systems
PACS-CS is a massively parallel cluster to be developed by the Project to carry out computational breakthroughs described above. The design targets are a peak performance of 14.3 Tflops with 2560 nodes connected by a 3-dimensional Hyper-Crossbar Network. The target for the start of operation is July 2006.
The development of the PACS-CS computer is being carried out through a close collaboration of the computer scientists and scientists of the target areas (materials and life sciences, particle and astrophysics, geoenvironmental and biological sciences). A working group containing members from the two areas are set up in each of the target areas, and regular meetings are held in which technical points for the hardware and software development of the PACS-CS are discussed and decided. Optimization of applications codes, for each node and its parallelization, is an important issue at present. The PACS-CS design details are being fixed step by step through these procedures.
For manufacturing of the PACS-CS computer, Hitachi Ltd. was selected on July 1st 2005 through a formal bidding process. For developing the Hyper-Crossbar Network driver software, Fujitsu Ltd. was selected on August 1st 2005. The Project members are working in close collaboration with these companies to develop the computer.