|Research Division on Supercomputing Systems, Cyberscience Center
In our group, collaboration with Takizawa laboratory and Kobayashi and Sato laboratory, we are working on research on high-performance computing systems supporting future progress in the computer science. Recent high-performance computing (HPC) systems are used in various fields and are being utilized not only for cutting-edge scientific research but also as a social infrastructure that supports our daily lives. On the other hand, since the HPC systems are getting larger and more complicated at a remarkable speed, making it difficult to exploit their potentials easily. Under this situation, we are also working on program optimization techniques to derive the performance of supercomputers and system software supporting them.
We are also working on designing energy-efficient computing systems from the viewpoint of architecture and hardware design. In future computing systems (incl. HPC systems), not only high performance but also extremely high energy efficiency is required, so we should drastically reduce their power consumption. Therefore, in the “More than Moore Era,” we should utilize innovative device technologies such as TSC (Through Silicon Via) and CNT (carbon nanotube), to support coming IoT and cloud-edge computing environments. We are working on the design and development of microprocessors and computing systems that realize the both of low power consumption and high performance beyond the technology scaling. For details, please see the research contents.
Since Moore’s law which has strongly supported the evolution of computer system so far is coming to an end, how to design, control, and utilizing computing systems are quite challenging issues.
We are working on overcoming these issues by hardware and software co-designs toward developing future computing systems, which support our future safety, secure, and healthy life.
High Performance Computing，Computing Architecture