Japanese analysis institute RIKEN is about to combine Nvidia GB200 NVL4 techniques into its two new supercomputers for scientific AI analysis and quantum computing.
The AI-dedicated system will utilise 1,600 Nvidia Blackwell graphics processing items (GPUs) and shall be related utilizing Nvidia Quantum-X800 InfiniBand networking.
Entry deeper business intelligence
Expertise unmatched readability with a single platform that mixes distinctive knowledge, AI, and human experience.
It goals to help analysis throughout life sciences, supplies science, local weather and climate forecasting, manufacturing, and laboratory automation.
The quantum computing system will incorporate 540 Nvidia Blackwell GPUs, additionally on the GB200 NVL4 platform and linked by Quantum-X800 InfiniBand networking.
Researchers are anticipated to make use of this infrastructure for work on quantum algorithms, hybrid simulation, and strategies that mix quantum and classical computing.
These developments construct on a collaboration introduced in August 2025, between Fujitsu and Nvidia to codesign a brand new flagship supercomputer beneath the event code title FugakuNEXT, which can succeed the present Fugaku supercomputer.
The 2 GPU-accelerated supercomputers at RIKEN will function proxy machines for {hardware}, software program, and software improvement referring to FugakuNEXT.
Operational launch of the 2 new supercomputers is scheduled for spring 2026, whereas FugakuNEXT is focused for deployment by 2030.
FugakuNEXT is deliberate to characteristic FUJITSU-MONAKA-X CPUs paired with Nvidia applied sciences by way of Nvidia NVLink Fusion, a newly launched silicon for high-bandwidth connectivity between Fujitsu CPUs and Nvidia GPU structure.
Nvidia famous that FugakuNEXT is projected to ship a hundredfold improve in software efficiency in comparison with present CPU-based or different current techniques, with plans to combine production-level quantum computer systems sooner or later.
Nvidia hyperscale and high-performance computing (HPC) vice chairman Ian Buck mentioned: “RIKEN has lengthy been one of many world’s nice scientific establishments, and immediately it stands on the forefront of a brand new period in computing.
“Collectively, we’re serving to Japan construct the muse for sovereign innovation that may drive breakthroughs to unravel the world’s most advanced scientific and industrial challenges.”
The mixed use of MONAKA-X CPUs and Nvidia’s newest GPUs is meant to facilitate developments throughout high-performance computing (HPC), AI, quantum analysis, and their intersections.
The initiative represents ongoing collaboration between RIKEN and Nvidia, with the businesses additionally working collectively on floating level emulation software program designed to leverage Nvidia Tensor Core GPU efficiency for conventional scientific computing.
This strategy is meant to allow purposes to totally utilise GPU energy for AI and HPC at RIKEN in addition to different supercomputing centres globally.
RIKEN Centre for Computational Science director Satoshi Matsuoka mentioned: “Integrating the Nvidia GB200 NVL4 accelerated computing platform with our next-generation supercomputers represents a pivotal development for Japan’s science infrastructure.”
RIKEN can also be set to undertake Nvidia CUDA-X, which incorporates over 400 libraries, microservices, and instruments optimised for GPU acceleration.
Not too long ago, Japan-based SoftBank Group bought all of its shares in Nvidia, producing $5.83bn from the transaction.
Join our day by day information round-up!
Give your online business an edge with our main business insights.
