, the pioneer in high performance artificial intelligence (AI) compute, today announced, for the first time ever, the simulation of a high-resolution natural convection workload at near real-time rates. Created using the WSE Field-equation API (WFA) developed by Department of Energy’s , the simulation, powered by NSF-funded Neocortex at the Pittsburgh Supercomputing Center (PSC), is expected to run several hundred times faster than what is possible on traditional distributed computers, as has been previously demonstrated with similar workloads.
The workload tested was thermally driven fluid flows, also known as natural convection, which is an application of computational fluid dynamics (CFD). Fluid flows occur naturally all around us — from windy days, to lake effect snowstorms, to tectonic plate motion. This simulation, made up of about 200 million cells, focuses on a phenomenon known as “Rayleigh-Bénard” convection, which occurs when a fluid is heated from the bottom and cooled from the top. In nature, this phenomenon can lead to severe weather events like downbursts, microbursts, and derechos. It’s also responsible for magma movement in the earth’s core and plasma movement in the sun.
A video including a visualization of the work can be seen at .
“We’re thrilled by the potential of this real-time natural convection simulation, as it means we can dramatically accelerate and improve the design process for some really big projects that are vital to mitigate climate change and enable a secure energy future — projects like carbon sequestration and blue hydrogen production,” said Dr. Brian J. Anderson, Lab Director at NETL. “This simulation wouldn’t have been possible without Cerebras’ industry-leading CS-2 system, which is the main compute element of PSC’s NSF-funded Neocortex supercomputer. Running on a conventional supercomputer, this workload is several hundred times slower, which eliminates the possibility of real-time rates or extremely high-resolution flows.”
In November 2022, Cerebras and NETL introduced a new field equation modeling API, powered by the , that was as much as 470 times faster than what was possible on NETL’s , delivering speeds beyond what either clusters of any number of CPUs or GPUs are currently able to achieve. Using a simple Python API that enables wafer-scale processing for much of computational science, WFA delivers gains in performance and usability that could not be obtained on conventional computers and supercomputers. The domain-specific, high-level programmer’s toolset WFA outperformed OpenFOAM® on NETL’s Joule 2.0 supercomputer by over two orders of magnitude in time to solution.
“We have a long-standing partnership with both NETL and PSC, and we’re proud of our continued innovation and advancements in scientific compute,” said Andrew Feldman, co-founder and CEO, Cerebras Systems. “This groundbreaking natural convection simulation delivers faster optimization at a lower cost, enabling for the first time an ultra-high fidelity digital twin running just ahead, predicting problems and suggesting improvements that will have substantial impact on the real world.”
The research was led by Dr. Dirk Van Essendelft, Machine Learning and Data Science Engineer at NETL, Terry Jordan, Computational Scientist; Dr. Mino Woo, Postdoctoral Research Fellow at NETL; Dr. Wei Shi, Research Engineer and Scientist at NETL; Robert Schreiber, Distinguished Engineer at Cerebras Systems; Michael James, co-founder and Chief Architect for Advanced Technologies at Cerebras Systems; Paola Buitrago, Neocortex Principal Investigator and PSC Director of Artificial Intelligence and Big Data; and Julian Uran, Machine Learning Research Engineer at PSC. Because of the simplicity of the WFA API, the results were achieved in just a few weeks and continue the close collaboration between NETL, PSC and Cerebras Systems.
“The NSF-funded Neocortex AI supercomputer was introduced mid-2020 as a technology testbed to explore interactive AI training for large deep learning models. The system is offered at no cost to the national research community,” said Paola Buitrago, Neocortex Principal Investigator and PSC Director of Artificial Intelligence and Big Data. “We are overjoyed by this recent achievement, in partnership with NETL and Cerebras Systems, as it demonstrates Neocortex’s potential to advance fundamental research via the acceleration of well-established algorithms that are complementary to AI.“
More information on this simulation can be found at .
About Cerebras Systems: is a team of pioneering computer architects, computer scientists, deep learning researchers, and engineers of all types. We have come together to build a new class of computer system, designed for the singular purpose of accelerating AI and changing the future of AI work forever. Our flagship product, the is powered by the world’s largest processor – the 850,000 core Cerebras enables customers to accelerate their deep learning work by orders of magnitude over graphics processing units.
About NETL: NETL is a U.S. Department of Energy national laboratory that drives innovation and delivers technological solutions for an environmentally sustainable and prosperous energy future. By leveraging its world-class talent and research facilities, NETL is ensuring affordable, abundant and reliable energy that drives a robust economy and national security, while developing technologies to manage carbon across the full life cycle, enabling environmental sustainability for all Americans.
About PSC: PSC’s mission is to enable the advancement of science and research. We cultivate collaborative partnerships, empower the next generation of researchers, and provide cutting-edge cyberinfrastructure. To learn more about Neocortex, please email [email protected], visit the , and/or subscribe to the .
Media Contacts:
RSS