NVIDIA-Powered Supercomputers Win Top Honors at Gordon Bell Prize, Pioneering Open Science
Groundbreaking research in climate modeling, tsunami forecasting, adn more, powered by NVIDIA’s cutting-edge supercomputing platforms, has been recognized with the prestigious Gordon Bell Prize.
Five finalists leveraging the alps,JUPITER,and Perlmutter supercomputers achieved remarkable breakthroughs in high-performance computing (HPC),with two teams ultimately claiming the top prize at SC25. These advancements, spanning climate science, materials science, fluid dynamics, and geophysics, demonstrate the transformative power of accelerated computing and open science principles.
The University of Texas at Austin, Lawrence Livermore National Laboratory, and the University of California San Diego were jointly awarded the Gordon Bell Prize for developing the world’s first digital twin capable of issuing real-time, probabilistic tsunami forecasts based on a full-physics model. This system, applied to the Cascadia subduction zone in the Pacific Northwest, achieved a staggering 10 billion-fold speedup – completing computations in 0.2 seconds that would traditionally take 50 years on 512 GPUs. “For the first time, real-time sensor data can be rapidly combined with full-physics modeling and uncertainty quantification to give people a chance to act before disaster strikes,” explained Omar Ghattas, a professor of mechanical engineering at UT Austin. “This framework provides a basis for predictive,physics-based emergency-response systems across various hazards.”
A second Gordon Bell Prize was awarded to the team behind the ICON Earth system model – a collaboration between the Max Planck Institute for Meteorology, German Climate Computing Center (DKRZ), Swiss National Supercomputing Centre (CSCS), Jülich Supercomputing Centre (JSC), ETH Zurich, the University of Hamburg, and NVIDIA. Their novel configuration allows for modeling the entire Earth’s systems at kilometer-scale resolution, capturing the flow of energy, water, and carbon with unprecedented detail. Running on the JUPITER supercomputer, ICON achieved a world record in global climate simulation, simulating 146 days every 24 hours. According to Daniel Klocke, a computational infrastructure and model development group leader at the Max Planck Institute fo
r Meteorology, “The ICON project is a prime example of how open science and collaborative efforts can lead to groundbreaking results. The NVIDIA platform was instrumental in achieving the performance and scalability required for this achievement.”
ORBIT-2: accelerating AI for Scientific Discovery: The Oak Ridge National Laboratory (ORNL) team developed ORBIT-2, a framework for accelerating scientific discovery using AI. ORBIT-2 leverages NVIDIA’s supercomputing technologies to analyze complex scientific data. The project achieved a 10x speedup compared to previous methods.The team used the Alps supercomputer. “NVIDIA’s advanced supercomputing technologies enabled ORBIT-2 to achieve exceptional scalability, reliability and impact at the intersection of AI and high-performance computing on NVIDIA platforms,” said Prasanna Balaprakash, director of AI programs at Oak Ridge National Laboratory.
QuaTrEx: Advancing Transistor Design through Nanoscale Device Modeling: Researchers at ETH Zurich have advanced nanoscale electronic device modeling with QuaTrEx, enabling faster and more accurate design of next-generation transistors. Running on the Alps supercomputer with NVIDIA GH200 Superchips, QuaTrEx can simulate devices with over 45,000 atoms.”Access to Alps was instrumental in the development of QuaTrEx,” noted Mathieu Luisier, a professor of computational nanoelectronics at ETH Zurich. “It allowed us to simulate devices that we could not imagine handling just a few months ago.” A simulation of electron flow in a nanoribbon transistor illustrates the project’s advancements.
Simulating Spacecraft at Record-Breaking Scales With the MFC Flow Solver: The Georgia Institute of Technology, in collaboration with NVIDIA, developed MFC, an open-source solver that enables fluid flow simulation 4x faster and with over 5x greater energy efficiency. A video showcasing a rocket engine simulation using computational fluid dynamics highlights the solver’s capabilities. “our new information geometric regularization method, combined with the NVIDIA GH200 Superchip’s unified virtual memory and mixed-precision capabilities, has drastically improved the efficiency of simulating complex computational fluid flows,” explained Spencer Bryngelson, an assistant professor at the Georgia Institute of Technology.
For the tsunami digital twin, ICON, and MFC projects, NVIDIA CUDA-X libraries were crucial for maximizing performance and efficiency. ICON also utilizes NVIDIA CUDA Graphs to streamline complex simulations.
Those interested in learning more about the latest supercomputing advancements can join NVIDIA at SC25, running through Thursday, November 20.
