Fujitsu Laboratories has received an order from the National Research and Development Agency, RIKEN for the basic design of the entire system, computational nodes, and CPU parts of the new flagship system, the next generation of the supercomputer “Fugaku.” The basic design period is until February 27, 2026.
With the rapid progress of technological innovation, including generative AI, the demand for computing resources required for research and development is expanding and diversifying rapidly . ) , the importance of efforts that combine AI with simulation, real-time data, and automated experiments (AI for Science) is increasing, and countries around the world are promoting the development and sophistication of their own computing infrastructure as a national strategy. In order for Japan to lead the world in science, technology and innovation and develop society and industry, it is essential to realize a new computing infrastructure that can flexibly respond to changes in demand, including AI for Science, and always meet the demands of the times. Against this background, the HPCI Project Promotion Committee decided to appoint RIKEN as the developer of a new flagship system, and development is underway.
Our company’s purpose is to “build trust in society through innovation and make the world more sustainable,” and we are committed to continuing research and development of technology and giving back to society, aiming to achieve both business transformation and the realization of a sustainable society. We view our participation in the basic design of “Fugaku NEXT” as one of the important initiatives to realize this purpose.
こちらもお読みください： DGXクラウドLepton、EUの開発者をNVIDIAネットワークに接続
富士通 will be carrying out the basic design of Fugaku NEXT, including the overall system, compute nodes, and CPU section. In addition to the knowledge and experience gained from pursuing performance through Fugaku, this basic design will incorporate the general-purpose CPU “FUJITSU-MONAKA ( Note 3 ) ” currently under development. ” currently under development. In addition to these technologies, Fugaku NEXT will aim to be a system that can flexibly respond to diverse changes in demand by linking with accelerators such as GPUs.
“FUJITSU-MONAKA” adopts cutting-edge 2-nanometer technology and applies Fujitsu’s proprietary technologies, such as microarchitecture optimized for the latest 3D package structure and ultra-low voltage circuit operation technology, to achieve both performance and power saving in a wide range of applications required for next-generation computing, from edge computing to data centers, and aims to provide value such as safety, security, and ease of use. The successor CPU to “FUJITSU-MONAKA” for “Fugaku NEXT”, “FUJITSU-MONAKA-X” (tentative name), will inherit and accelerate the application assets developed for “Fugaku”, and will meet the demand for AI processing by incorporating cutting-edge AI processing acceleration functions. This CPU is not only intended to be installed on “Fugaku NEXT”, but is also expected to be used in a wide range of fields that support society and industry.
We will continue to provide innovation and reliability through our domestically produced CPU technology, which is at the core of our world-leading computing infrastructure, in order to contribute to the development of science and technology in Japan and to solving problems in business and society. Furthermore, through these developments, we will promote collaboration with various companies, researchers, and communities, expand the ecosystem for research and development of applications, AI, etc., and contribute to the realization of a sustainable society.
ソース 富士通
