[Disclaimer] This article is reconstructed based on information from external sources. Please verify the original source before referring to this content.
News Summary
The following content was published online. A translated summary is presented below. See the source for details.
RIKEN, Japan’s leading research institute, has partnered with Fujitsu and NVIDIA to develop Japan’s next flagship supercomputer, FugakuNEXT. The project, launched in August 2023, aims to create a landmark high-performance computing system operational by 2030. Unlike its predecessor Fugaku, FugakuNEXT will incorporate NVIDIA GPUs, marking a significant evolution in Japanese supercomputing. The system targets performance exceeding 600 exaFLOPS in FP8 precision, aspiring to become the world’s first “zetta-scale” supercomputer. FugakuNEXT is designed to address Japan’s pressing scientific needs, focusing on the convergence of AI and high-performance computing. It will support various applications, including AI for science, advanced manufacturing, and quantum computing integration. The project, with a budget exceeding $750 million, represents a major step in Japan’s commitment to maintaining its position at the forefront of global supercomputing capabilities.
Source: NVIDIA
Our Commentary
Background and Context
The FugakuNEXT project represents a significant leap in supercomputing capabilities for Japan. It builds upon the success of its predecessor, Fugaku, which was completed in 2021. The collaboration between RIKEN, Fujitsu, and NVIDIA marks a strategic shift in Japan’s approach to high-performance computing, integrating GPUs into its flagship supercomputer for the first time to enhance AI and simulation capabilities.
Expert Analysis
The development of FugakuNEXT signifies Japan’s commitment to maintaining its competitive edge in the global supercomputing landscape. By targeting zetta-scale performance, the project aims to push the boundaries of what’s possible in scientific computing and AI research.
Key points:
- Integration of NVIDIA GPUs represents a major architectural shift from CPU-only systems
- Focus on AI-HPC convergence aligns with emerging trends in scientific computing
- Ambitious performance goals set new benchmarks for supercomputing capabilities
Additional Data and Fact Reinforcement
The FugakuNEXT project is characterized by several ambitious targets and features:
- Target performance: Over 600 exaFLOPS in FP8 precision
- Expected to be 5-10 times more powerful than Fugaku
- Project budget exceeds $750 million
- Planned operational date: Around 2030
Related News
The development of FugakuNEXT occurs against a backdrop of global competition in supercomputing, with countries like the United States and China also investing heavily in next-generation systems. This project aligns with Japan’s broader strategy to maintain its technological leadership and support cutting-edge research across various scientific domains.
Summary
The FugakuNEXT project represents a significant evolution in Japan’s supercomputing capabilities, combining traditional high-performance computing with advanced AI technologies. Its success could redefine the landscape of scientific research and industrial innovation, positioning Japan at the forefront of the next generation of computational science.
