RIKEN taps Fujitsu and NVIDIA for FugakuNEXT supercomputer focused on Artificial Intelligence

RIKEN has launched an international collaboration with Fujitsu and NVIDIA to co-design FugakuNEXT, the successor to Fugaku, pairing simulation with Artificial Intelligence workloads. The early award of the contract lets the partners shape the system to address Japan´s scientific priorities and industrial needs.

RIKEN announced the start of an international collaboration with Fujitsu and NVIDIA to co-design FugakuNEXT, the successor to the supercomputer Fugaku. The initiative was unveiled at the FugakuNEXT International Initiative Launch Ceremony in Tokyo on Aug. 22, where RIKEN president Makoto Gonokami and Satoshi Matsuoka, director of the RIKEN Center for Computational Science, outlined the project. Fujitsu chief technology officer Vivek Mahajan and Ian Buck, vice president of hyperscale and high-performance computing at NVIDIA, attended and stressed the collaborative design approach. The contract was awarded early so the three partners can work side by side to shape the system´s architecture around Japan´s most urgent research priorities.

FugakuNEXT is being designed as a hybrid Artificial Intelligence and high-performance computing system that combines simulation and Artificial Intelligence workloads. The design will feature FUJITSU-MONAKA-X CPUs paired with NVIDIA technologies through NVLink Fusion, a new silicon approach for high-bandwidth connections between Fujitsu processors and NVIDIA architecture. The project will draw on NVIDIA´s software stack, including NVIDIA CUDA-X libraries such as NVIDIA cuQuantum for quantum simulation, RAPIDS for data science, NVIDIA TensorRT for inference and NVIDIA NeMo for large language model development, alongside domain-specific software development kits tailored for science and industry. RIKEN, Fujitsu and NVIDIA plan coordinated software work on mixed-precision computing, continuous benchmarking and performance optimization.

The teams expect FugakuNEXT to serve a broad set of applications, from automating hypothesis generation, code creation and experiment simulation to accelerating scientific research with surrogate models and physics-informed neural networks. Target domains named in the announcement include earth systems modeling and disaster preparedness, drug discovery and advanced manufacturing, where AI-trained models can learn from simulations to speed and improve design. Backed by Japan´s MEXT, FugakuNEXT is positioned as a sovereign infrastructure to support universities, government agencies and industry partners nationwide and to showcase Japanese innovations that could become blueprints for the world.

82

Impact Score

Saudi Artificial Intelligence startup launches Arabic LLM

Misraj Artificial Intelligence unveiled Kawn, an Arabic large language model, at AWS re:Invent and launched Workforces, a platform for creating and managing Artificial Intelligence agents for enterprises and public institutions.

Introducing Mistral 3: open artificial intelligence models

Mistral 3 is a family of open, multimodal and multilingual Artificial Intelligence models that includes three Ministral edge models and a sparse Mistral Large 3 trained with 41B active and 675B total parameters, released under the Apache 2.0 license.

NVIDIA and Mistral Artificial Intelligence partner to accelerate new family of open models

NVIDIA and Mistral Artificial Intelligence announced a partnership to optimize the Mistral 3 family of open-source multilingual, multimodal models across NVIDIA supercomputing and edge platforms. The collaboration highlights Mistral Large 3, a mixture-of-experts model designed to improve efficiency and accuracy for enterprise artificial intelligence deployments starting Tuesday, Dec. 2.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.