Masayuki Umemura, Professor, Director of CCS, University of Tsukuba
Taisuke Boku, Professor, Deputy Director of CCS, University of Tsukuba
Yoshinobu Kuramashi, Professor, CCS, University of Tsukuba
CCS, University of Tsukuba, Japan
The Center for Computational Sciences (CCS) at University of Tsukuba was originally established as the Center for Computational Physics (CCP) in 1992 as a research center focusing on a collaborative research among computational physicists and computer scientists toward world top level computational physics as well as high performance computing. After 12 years of research activity, the center was reorganized as Center for Computational Sciences in 2004 with expanded research area on domain science and computer science. Currently more than 30 of faculty staffs belong to the center for collaborative research on domain science including particle physics, astrophysics, nuclear physics, life science and environmental science, and computer science including high performance computing and informatics. The mission of the CCS is to promote scientific discovery by computational science through the application of advanced computing technologies, and to support researches of computational science in Japanese universities and institutes by operating leading-edge computing systems.
Collaborative Interdisciplinary Program
As a nation-wide collaborative research center, the CCS operates a special program for supercomputer resource sharing named Collaborative Interdisciplinary Program to provide a large capacity of supercomputer utilization time to the research on wide area of computational sciences. The center accepts scientific research projects every year to provide the supercomputer managed by the center with free of charge to promote wide variety of advanced computational science and high performance computing researches. In every year, more than 40 of projects are selected in this program, and hundreds of users perform their advanced research based on large scale simulation.
Joint Center for Advanced High Performance Computing (JCAHPC)
In 2013, the Center for Computational Sciences, University of Tsukuba (CCS) and the Information Technology Center, the University of Tokyo agreed to establish the Joint Center for Advanced High Performance Computing (JCAHPC). Primary mission of JCAHPC is designing, installing and operating the Post T2K System based on many-core architectures, such as Intel® Xeon Phi™. The Post T2K System is expected to be 20-30 PFLOPS of peak performance, and will be installed in FY 2015. CCS and ITC will develop system software, numerical libraries, and large-scale applications to for the Post T2K system under intensive collaboration through JCAHPC.
One of the most historical and primary research topics in the CCS is that on particle physics field, Quantum Chromo-Dynamics (QCD). This is one of the most fundamental physics problems where the origin of materials, the elementary particles and quark are analyzed on their characteristics and behavior. Through the history of the center, we have been developing the primary simulation code on QCD with large scale parallelism which can be executed on various supercomputers from PC clusters to K Computer. The core simulation code in our research is summarized as a benchmark code QCDBench, which has been used for benchmarking on all the supercomputer procurement and evaluation through the history of the CCS. Currently, we are focusing on the development of optimized code of QCDBench for many-core architecture processor Intel Xeon Phi.
- Taisuke Boku, Ken-Ichi Ishikawa, Yoshinobu Kuramashi, Lawrence Meadows, Michael D'Mello, Maurice Troute, Ravi Vemuri, December 2016, A performance evaluation of CCS QCD Benchmark on the COMA (Intel(R) Xeon PhiTM, KNC) system, arXiv.org
- Larry Meadows, Ken-Ichi Ishikawa, Taisuke Boku and Masashi Horikoshi, January 2018, Multiple Endpoints for Improved MPI Performance on a Lattice QCD Code, IXPUG Workshop Asia 2018
- Masahiro Nakao, Hitoshi Murai, Taisuke Boku, Mitsuhisa Sato, January 2018, Performance Evaluation for Omni XcalableMP Compiler on Many-core Cluster System based on Knights Landing, IXPUG Workshop Asia 2018
- Masashi Horikoshi, Larry Meadows, Thomas Elken, Pradeep Sivakumar, Edward Mascarenhas, James Erwin, Dmitry Durnov, Alexander Sannikov, Alexey Malhanov, Toshihiro Hanawa and Taisuke Boku, January 2018, Scaling Collectives on Large Clusters of Intel® Architecture Processors, IXPUG Workshop Asia 2018