Cerebras Systems, in collaboration with researchers from Sandia, Lawrence Livermore, and Los Alamos National Laboratories, has set a new world record in molecular dynamics simulations. Using a single Cerebras Wafer Scale Engine, researchers achieved over 1 million simulation steps per second. This performance is 748 times faster than what is achievable on the world’s leading supercomputer ‘Frontier.’
The Cerebras system also outperformed Anton 3, the world’s leading purpose-built supercomputer for molecular dynamics.
While Anton 3 uses 512 specialized processors and 400 kilowatts of power, the Cerebras system uses a single accelerator, consumes only 7% of the power, and outperforms Anton 3 by 20%. Michael James, Chief Architect of Advanced Technologies and co-founder of Cerebras Systems, said, “This new world record means that scientists can now complete two years’ worth of GPU-based simulation work every single day. This will greatly accelerate the rate of innovation derived from molecular simulations.
This breakthrough will provide insights into material structure and function and, when extended to biomolecules, unlock new capabilities in protein folding, medicine, and drug development.
Traditionally, supercomputing performance is achieved via weak scaling, which increases the size of a simulation, or strong scaling, which increases its speed. Existing supercomputer designs use many GPUs to scale simulation size at the same speed. In contrast, Cerebras achieves strong scaling by accelerating simulation speed hundreds of times over today’s leading supercomputers.
Cerebras outpaces supercomputers in simulations
This allows scientists to simulate materials longer and effectively view their future states. James H.
Laros III, Distinguished Member of Technical Staff and AMT program lead, stated, “The partnership between the NNSA laboratories and Cerebras Systems is part of the Advanced Memory Technology (AMT) program, which aims to accelerate exascale supercomputers by 40x as early as 2025. With Cerebras’ currently deployed wafer-scale computers, the teams achieved this materials science breakthrough and a speedup that exceeded the goal of the AMT program by more than 4 times. This experience bodes well for future impacts on our program and potential scientific advances.
Long timescale simulations now allow scientists to explore previously inaccessible phenomena across a wide range of domains.
Materials scientists can study the long-term behavior of complex materials, such as the evolution of grain boundaries in metals, leading to the development of more resilient materials. Pharmaceutical researchers can simulate protein folding and drug-target interactions over physiologically relevant timescales, accelerating the discovery of life-saving therapies. Renewable energy experts can optimize catalytic reactions and design more efficient energy storage systems by simulating atomic-scale processes over extended durations.
This work has been nominated as a finalist for the ACM Gordon Bell Prize.
April Isaacs is a news contributor for DevX.com She is long-term, self-proclaimed nerd. She loves all things tech and computers and still has her first Dreamcast system. It is lovingly named Joni, after Joni Mitchell.
