Super Parallel Computing Project: Scientific Research on Priority Areas for Ministry of Education

Hidehiko TANAKA (tanaka@mtl.t.u-tokyo.ac.jp)

Department of Electrical Engineering
The University of Tokyo, Tokyo 113, JAPAN


Abstract

As one of the scientific research on priority areas for the Ministry of Education, we started a project of super parallel computing from 1992. This is a basic research of super parallel computing expected to be the key technology for the science and technology in the future. For this project, we organized 5 groups of university researchers. These groups, A, B, C, D and Control take charge of applications and modeling, parallel programming languages, operating systems, hardware system, and management for the 4 groups mentioned before, respectively. As our major target is to show the feasibility of super parallel computing in the region where the number of processors are more than 10,000, through building a system prototype for super parallel computing, almost all of the research themes are formed around this prototype. For example, Group A has a role of proposing an evaluation set of applications for super parallel computing. Group B are designing and implementing a language NCX based on language C. Group C are designing and building an operating system based on the micro kernel. Group D are designing and implementing a hardware prototype made of 1024 processing elements. The duration of this project is 3 years. It is now at the most busy time for the design of this prototype. In this presentation, the details of this project will be discussed.