Nvidia Joins Race to Build Fugaku's Successor

Nvidia Joins Effort to Develop Next-Generation Supercomputer

Nvidia, a leading U.S. semiconductor company, has announced its involvement in a new initiative aimed at creating a successor to Japan's Fugaku supercomputer. This project, known as "FugakuNEXT," is being developed by Riken, a state-backed research institute in Japan. The collaboration includes Fujitsu Ltd., which will contribute central processing units (CPUs), and Nvidia, which will provide graphics processing units (GPUs) to support advanced computing for artificial intelligence.

The goal of the FugakuNEXT project is to create a system that offers world-leading capabilities in high-performance computing. The supercomputer is expected to begin operations around 2030, marking a significant step forward in computational power and efficiency. By integrating Nvidia’s GPUs with Fujitsu’s CPUs, the system is anticipated to deliver unprecedented speed and performance, making it a powerful tool for various scientific and technological applications.

Potential Applications and Capabilities

FugakuNEXT is expected to have a wide range of applications across multiple industries. One of the primary areas where this supercomputer could make a significant impact is in drug development. Its advanced computing capabilities could accelerate the discovery of new medicines and treatments, potentially reducing the time required for clinical trials and improving patient outcomes.

In addition, the supercomputer could revolutionize automobile design by enabling more complex simulations and testing of vehicle performance, safety, and efficiency. It could also play a crucial role in addressing climate change by supporting advanced modeling of environmental systems, helping scientists better understand and predict climate patterns.

The new system is projected to be five to ten times more powerful than its predecessor, Fugaku. It will be capable of performing up to one sextillion calculations per second, making it an essential tool for developing next-generation artificial intelligence applications. This level of computational power will enable researchers to tackle some of the most complex challenges in science and technology.

Legacy of Fugaku

Fugaku, named after an alternative term for Mount Fuji, was once the fastest supercomputer in the world. It held the top position in the TOP500 ranking of supercomputers in terms of computing speed in 2020. However, it lost that title in 2022 and currently ranks seventh in the global list. Despite this, Fugaku has made significant contributions to scientific research and remains an important milestone in the history of high-performance computing.

FugakuNEXT will be located in Kobe, Japan, the same site as its predecessor. This strategic location allows for continuity in research and development while leveraging existing infrastructure and expertise in the region. The decision to build the new supercomputer in the same location highlights the importance of maintaining a strong research presence in western Japan.

Future Outlook

With the involvement of industry leaders like Nvidia and Fujitsu, FugakuNEXT represents a major advancement in the field of supercomputing. The collaboration between these companies and Riken underscores the growing importance of international partnerships in driving technological innovation. As the demand for high-performance computing continues to grow, projects like FugakuNEXT will play a vital role in shaping the future of science, technology, and industry.