RIKEN and Argonne National Lab Partner on US-Japan AI Supercomputing Collaboration with Nvidia and Fujitsu

Pioneering Next-Gen AI Research Through Trans-Pacific HPC Alliance

New0 comments

Be one of the first to share your thoughts!

Add your comments now!

Have your say

Engagement level

See more Higher Ed News Articles

the number forty forty forty forty forty forty forty forty forty forty forty forty forty forty
Photo by BoliviaInteligente on Unsplash

RIKEN and Argonne Seal Landmark MoU for AI-Driven Supercomputing Innovation

The RIKEN Center for Computational Science (R-CCS) in Japan has signed a pivotal Memorandum of Understanding (MoU) with the U.S. Department of Energy's Argonne National Laboratory (ANL), Fujitsu Limited, and NVIDIA Corporation. Announced on January 27, 2026, this collaboration marks a significant step in US-Japan AI supercomputing efforts, focusing on developing next-generation infrastructure to tackle complex scientific challenges. 28 37

This partnership builds on prior agreements, including the 2024 MoU between RIKEN and ANL on AI for Science, aiming to accelerate joint research in high-performance computing (HPC) architectures, system software, and AI applications. RIKEN's expertise with the Fugaku supercomputer, once the world's fastest, complements Argonne's Aurora exascale system, promising breakthroughs in fields like climate modeling, drug discovery, and materials science. 80

Key Players and Their Roles in the Collaboration

RIKEN R-CCS, located in Kobe, Japan, leads national supercomputing initiatives and hosts Fugaku, which has driven thousands of research publications since 2021. Argonne National Laboratory, managed by UChicago Argonne LLC for the DOE, operates Aurora, one of the first exascale supercomputers capable of over one exaFLOP performance, revolutionizing AI workloads. 67

Fujitsu brings its A64FX ARM-based processors from Fugaku and forthcoming FUJITSU-MONAKA-X CPUs for FugakuNEXT, while NVIDIA contributes Blackwell GPUs and NVLink Fusion technology for seamless CPU-GPU integration. This quartet's synergy positions them to pioneer zettascale computing by 2030. 61

In Japan, RIKEN collaborates extensively with universities like the University of Tokyo, Kyoto University, and Tohoku University, fostering researcher exchanges and joint projects. This MoU opens doors for university faculty and students to access cutting-edge resources, boosting publications in top journals. 92

Evolution from Prior US-Japan AI Partnerships

Prior to this MoU, RIKEN and Argonne collaborated on AuroraGPT, training large language models on scientific datasets as part of the Trillion Parameter Consortium. Fugaku has topped Graph500 rankings 11 times, yielding highlights like COVID-19 aerosol simulations and protein folding studies, with over 100 press releases on research outputs in 2025 alone. 82 97

Recent NVIDIA-RIKEN announcements in November 2025 introduced two new systems: an AI supercomputer with 1,600 Blackwell GPUs and a quantum-HPC hybrid with 540 GPUs, precursors to FugakuNEXT. These integrate with Argonne's efforts under the DOE's Genesis Mission, launched by President Trump in November 2025 to harness AI for federal scientific data analysis. 5 47

Fugaku and Aurora supercomputers representing US-Japan AI collaboration

Technical Foundations: Supercomputers Powering the Future

FugakuNEXT aims for 100x performance gains over current systems, combining Fujitsu's MONAKA-X CPUs with NVIDIA's Grace Blackwell Superchips via NVLink Fusion. Argonne's Aurora, powered by Intel Xeon Max CPUs and Data Center GPUs, excels in AI training, processing petabytes of data for simulations unattainable elsewhere.

The collaboration will focus on:

  • Co-designing system software for mixed-precision computing and continuous benchmarking.
  • Developing AI applications for earth systems modeling and quantum algorithm validation. 62
  • Optimizing hybrid quantum-HPC workflows, as in RIKEN's ongoing quantum integration project. 90

Argonne researchers will showcase advancements at SCA/HPCAsia 2026, a joint US-Japan conference, highlighting data compression and big data techniques from their partnership. 79

Genesis Mission Integration

Trump's Genesis Mission allocates $320 million to AI platforms across DOE labs, including Argonne, for biotech, manufacturing, and quantum science. Japan's involvement aligns with this 'Manhattan Project-scale' effort, enhancing bilateral data sharing and model training. 118

RIKEN Official MoU Announcement

Anticipated Research Outputs and Publications

As a research publication news milestone, this partnership promises a surge in high-impact papers. Fugaku alone produced breakthroughs like Fugaku-LLM, a large language model for science, featured in 2025 highlights. Joint RIKEN-Argonne efforts could yield papers in Nature, Science, and specialized journals on lossy compression for scientific data. 105 83

Expect publications on:

  • Generative AI for materials discovery.
  • Scalable HPC architectures for climate prediction.
  • Quantum error correction via hybrid systems.

RIKEN's alliances with universities ensure co-authored works from academic teams, elevating Japan's global research profile.

Boost for Japanese Higher Education and Researchers

For Japan's higher education sector, this collaboration means expanded access to world-class HPC for university researchers. Institutions like Keio University and Yonsei (via RIKEN ties) gain from hybrid quantum platforms, while Tsukuba and Tokyo universities participate in Fugaku projects. 89 95

Opportunities include postdoc positions, faculty exchanges, and training in AI-HPC. Researchers can apply for research jobs leveraging these resources, enhancing CVs with international publications. This aligns with US-Japan university partnerships like University of Washington-Tsukuba with NVIDIA. 107

Statistics show Fugaku enabled 5,000+ projects in 2025, many university-led, predicting a 20-30% rise in AI-related PhDs and publications in Japan.

Japanese and US researchers collaborating on AI supercomputing

Career Implications and Actionable Insights for Academics

Aspiring AI-HPC experts in Japan should:

  • Pursue postdoc opportunities at RIKEN or partner labs.
  • Develop skills in NVIDIA CUDA and Fujitsu tools via online courses.
  • Collaborate on Fugaku proposals through HPCI for publication credits.
  • Explore academic CV tips for international applications.

Universities benefit from talent attraction; check university jobs for lecturer roles in computational science.

Argonne-RIKEN 2024 MoU Details

Challenges, Solutions, and Future Outlook

Challenges include data interoperability and energy efficiency for zettascale systems. Solutions: standardized software stacks and green computing benchmarks from joint R&D.

By 2030, expect FugakuNEXT online, powering exascale AI discoveries. This bolsters Japan's research sovereignty amid global AI race, with Argonne's Genesis integration fostering balanced US-Japan leadership.

For researchers, rate professors on Rate My Professor or seek higher ed jobs to join the wave. Stay informed via career advice.

a man sitting at a table using a laptop computer

Photo by Christopher Lee on Unsplash

Discussion

0 comments from the academic community

Sort by:
You

Please keep comments respectful and on-topic.

Frequently Asked Questions

🤝What is the RIKEN-Argonne AI supercomputing MoU?

Signed Jan 27, 2026, it advances joint AI-HPC research, building on 2024 agreement. Focus: next-gen infrastructure, software, apps.28

🚀How does FugakuNEXT fit into this collaboration?

FugakuNEXT, RIKEN's zettascale successor with Fujitsu CPUs & NVIDIA GPUs, integrates via MoU for co-development.

🔬What research areas will benefit?

AI for science: climate, drugs, materials. Expect joint publications from hybrid quantum-HPC.

🎓Impact on Japanese universities?

Enhanced access for Tokyo U, Kyoto U researchers; more postdocs, publications. Research jobs surge.

Role of Argonne's Aurora?

Exascale power for AI training; ties to DOE Genesis Mission for data-driven discovery.

💻NVIDIA and Fujitsu contributions?

Blackwell GPUs, NVLink; A64FX/MONAKA-X CPUs for seamless hybrid computing.

📚Past achievements from similar ties?

AuroraGPT LLMs, Fugaku Graph500 wins, COVID sims with 1000+ pubs.

💼Career opportunities?

Postdocs, faculty roles in AI-HPC. Check Rate My Professor & jobs.

🇺🇸Genesis Mission connection?

Trump's DOE initiative aligns, funding AI science with Argonne leadership.

📅Future timeline?

New systems 2026; FugakuNEXT 2030. More SCA/HPCAsia showcases.

🛤️How to get involved?

Apply HPCI proposals, RIKEN alliances. Career advice for academics.