NVIDIA and RIKEN are pushing the boundaries of scientific research in Japan with the introduction of two powerful supercomputers designed for AI and quantum computing. These systems, powered by NVIDIA's cutting-edge technology, will revolutionize the way Japan approaches scientific advancements and technological leadership.
The first supercomputer, named GB200 NVL4, will be equipped with 1,600 NVIDIA Blackwell GPUs, interconnected via NVIDIA Quantum-X800 InfiniBand networking. This setup will significantly enhance research in critical fields such as life sciences, materials science, climate and weather forecasting, manufacturing, and laboratory automation. By leveraging the GB200 platform, RIKEN aims to accelerate scientific breakthroughs and foster innovation in these sectors.
The second supercomputer, dedicated to quantum computing, will feature 540 NVIDIA Blackwell GPUs, also utilizing the GB200 NVL4 platform and NVIDIA Quantum-X800 InfiniBand networking. This system will play a pivotal role in advancing quantum algorithms, hybrid simulation, and quantum-classical computing methods, positioning Japan at the forefront of quantum research.
NVIDIA's partnership with RIKEN extends beyond these supercomputers. The collaboration includes the development of a flagship supercomputer, FugakuNEXT, in partnership with Fujitsu. FugakuNEXT will be an extraordinary leap forward in supercomputing, delivering 100 times greater application performance compared to traditional systems. It will be equipped with FUJITSU-MONAKA-X CPUs, paired with NVIDIA technologies using NVLink Fusion, enabling high-bandwidth connections between Fujitsu's CPUs and NVIDIA's architecture.
Furthermore, NVIDIA is working with RIKEN to develop floating-point emulation software, harnessing the power of NVIDIA Tensor Core GPUs to accelerate traditional scientific computing. This technology will unlock the full potential of GPUs for AI and HPC, benefiting RIKEN and supercomputing centers worldwide.
RIKEN also plans to leverage NVIDIA CUDA-X, a suite of GPU-accelerated libraries and tools, to enhance its cutting-edge HPC applications. This will further boost AI for science and quantum computing initiatives in Japan, solidifying the country's position as a leader in these fields.
The two new supercomputers will be operational in spring 2026, while FugakuNEXT is expected to be ready by 2030. These advancements in supercomputing software and hardware will undoubtedly unlock new scientific discoveries and drive Japan's innovation in AI, quantum computing, and high-performance computing.
