Argonne, RIKEN, Fujitsu, and NVIDIA partner to push AI + HPC for faster scientific discovery
Argonne National Laboratory has entered into a partnership with RIKEN, Fujitsu Limited and NVIDIA to advance artificial intelligence and high performance computing for science. The agreement, signed January 27, supports the U.S. Department of Energy's Genesis Mission to accelerate discovery science, strengthen national security and drive energy innovation.
Building on an earlier 2024 agreement between Argonne and RIKEN, the four organizations will collaborate on next-generation computing infrastructure, system software and applications for science and engineering, and the practical use of advanced AI methods across research domains.
According to Laboratory Director Paul Kearns, the collaboration is a pivotal step to apply AI and HPC to urgent challenges in energy, security and fundamental research. The partners plan to lay the groundwork for next-gen computing architectures and AI-driven discovery that support Genesis Mission goals.
What the collaboration covers
- Future HPC and AI architectures: Design and prototype systems that integrate modeling, simulation and AI to tackle complex scientific problems.
- Integrated HPC systems and platforms: Develop and validate end-to-end systems that bring hardware, software and facilities together as a seamless platform.
- Shared software ecosystem for AI-enabled science: Build a shared, open software stack for large-scale, AI-driven research.
- Flagship AI for science applications: Deliver high-impact scientific applications that demonstrate measurable outcomes and inform future systems and software.
- Laboratory experiment automation: Use AI and robotics to automate experiments and shorten data-to-insight cycles.
- Accelerated quantum computing: Explore combined AI and quantum approaches for practical, large-scale scientific workloads.
How this helps research teams
- Early access to prototype architectures can guide method development, scaling studies and workflow design.
- A shared software stack reduces duplicated effort and improves portability across facilities.
- Flagship applications provide benchmarks and patterns teams can reuse for performance and reliability.
- Experiment automation tightens feedback loops between design, execution and analysis.
Community activity
The partnership will run joint meetings, workshops and engagement with the broader community to accelerate practical results and share implementation lessons across labs and universities.
At the signing
At the signing ceremony, from left to right: RIKEN R-CCS Director Satoshi Matsuoka; Minister of Education, Culture, Sports, Science and Technology Yasuyoshi Kakita; Corporate Executive Officer, Corporate Vice President, CTO, in charge of System Platform, Fujitsu Limited Vivek Mahajan; RIKEN President Makoto Gonokami; Argonne Associate Laboratory Director for Computing, Environment and Life Sciences Rick Stevens; Vice President, Global Sales and Business Development for HPC/AI/Supercomputing, NVIDIA Corporation John Josephakis; DOE Under Secretary for Science Darío Gil. (Credit: RIKEN)
Learn more
If your lab is building AI skills for scientific workflows, see curated programs by role: AI courses by job.
Your membership also unlocks:
