New York City, U.S. – November 20, 2025 — Five cutting-edge research teams powered by NVIDIA-accelerated supercomputers have been named finalists for the prestigious 2025 Gordon Bell Prize, awarded for outstanding achievements in high-performance computing (HPC). Their work spans climate modeling, electronics design, fluid dynamics, geophysics, and real-time disaster simulation — all enabled by NVIDIA’s Grace Hopper superchips on Europe’s most powerful compute platforms. NVIDIA Blog
Revolutionary Science, Accelerated by NVIDIA HPC
The 2025 Gordon Bell finalists are pushing open science boundaries with projects running on three top-tier supercomputers:
- Alps, at the Swiss National Supercomputing Centre (CSCS), powered by more than 10,000 GH200 Grace Hopper superchips.
- JUPITER, Europe’s first exascale system, hosted at the Jülich Supercomputing Centre, leveraging Grace Hopper architecture and Quantum-X800 InfiniBand networking.
- Perlmutter, at the U.S. National Energy Research Scientific Computing Center (NERSC), running NVIDIA-accelerated compute.
These powerful systems are helping researchers accelerate discovery in fields with high societal value, such as forecasting extreme weather, designing next-generation semiconductors, simulating fluid flow, and even building early-warning systems for natural disasters.
Spotlight on the Finalist Projects
Here are the five standout projects named Gordon Bell finalists:
1. ICON – Earth System Modeling at Kilometer Scale
A consortium led by the Max Planck Institute, DKRZ, ETH Zurich, CSCS, Jülich, and NVIDIA is using the ICON model to simulate the Earth’s atmosphere, oceans, and land in unprecedented detail. Their kilometer-scale resolution allows them to simulate 146 days of Earth processes in just 24 hours — a breakthrough for climate science prediction and energy-system planning.
2. ORBIT-2 – AI Model for High-Resolution Climate Prediction
Developed by Oak Ridge National Laboratory and NVIDIA, ORBIT-2 is an exascale AI foundation model for weather downscaling. Running on the Alps supercomputer, it enables ultra-high-resolution local climate forecasts, helping to predict complex phenomena like urban heat islands, extreme rainfall, and micro-climate shifts.
3. QuaTrEx – Nanoscale Transistor Design
ETH Zurich researchers used QuaTrEx algorithms on the Alps supercomputer to simulate transistors at the atomic scale, modeling more than 45,000 atoms. This capability accelerates the design of next-generation semiconductors, addressing both performance and energy-efficiency challenges.
4. MFC Flow Solver – Spacecraft Fluid Dynamics
A team from Georgia Institute of Technology, in collaboration with NVIDIA, developed an open-source solver named MFC to simulate the fluid dynamics of rocket engine plumes. Running on Alps, their solver achieved a 4× speed-up and more than 5× energy efficiency over previous approaches — enabling design of more efficient, safer spacecraft.
5. Tsunami Digital Twin – Real-time Forecasting
Researchers from UT Austin, Lawrence Livermore National Lab, and UC San Diego created a digital twin for real-time tsunami forecasting. The twin combines live sensor data with full-physics simulations, running on Alps and Perlmutter to produce probabilistic tsunami predictions in milliseconds — a dramatic leap in early-warning potential.
Strategic Significance & Industry Impact
For Ixoraly’s community of technologists, investors, and strategy leaders, these developments carry meaningful implications:
- Open Science as a Growth Lever: These projects highlight how HPC + AI infrastructures are no longer exclusive tools — they are becoming engines of open scientific innovation.
- Sovereign & Planetary Risk Use Cases: From climate downscaling to disaster forecasting, the highlighted research addresses critical global challenges — with real-world impact potential.
- Semiconductor Innovation: QuaTrEx’s atomic-scale simulations could accelerate the next generation of chips, helping nations and firms leapfrog current limitations.
- HPC-AI Scaling: These achievements showcase how NVIDIA’s superchips and high-throughput interconnects enable scaling AI and physics-based compute to exascale levels.
- Resilience & Safety: The tsunami digital twin demonstrates how real-time physics models can become integrated into public-safety infrastructure, opening advisory and risk-management markets.
Next Steps & What to Watch
With the finalists now announced, attention turns to SC25 — the International Conference for High Performance Computing, Networking, Storage, and Analysis — where winners will be revealed. SC25 Key signals for Ixoraly’s audience to monitor include:
- Winner Announcement: The Gordon Bell Prize winners will be declared during SC25, shedding light on which projects deliver the most impactful compute innovations.
- Adoption & Scaling: Whether these finalist teams translate their proofs-of-concept into operational tools or commercial / mission-critical systems in climate science, aerospace, or semiconductors.
- Funding & Partnerships: New collaborations, research grants, or commercial spin-outs stemming from finalist projects will signal broader momentum.
- Ecosystem Growth: More institutions and countries may allocate resources to NVIDIA-powered HPC, accelerating the global adoption of HPC for science.
- Policy & Investment: Governments and investors may take note — such technological breakthroughs could drive funding in sovereign computing, sustainability, and next-gen hardware infrastructure.
About Ixoraly
Ixoraly is a global press release distribution platform dedicated to helping organizations share their stories with the world. Focused exclusively on corporate, business, financial, and industry news, Ixoraly enables companies to reach journalists, media outlets, investors, analysts, and digital audiences across international markets.
