Researchers at Georgia Tech have achieved a world-record resolution in turbulence simulations using the Frontier Exascale computer, opening new avenues for scientific discovery and innovation in various fields.
Researchers at Georgia Tech have made a groundbreaking advancement in the field of turbulence simulation, achieving a world-record resolution thanks to the unprecedented power of Frontier, the world’s fastest Exascale computer. This development promises to deepen our understanding of complex phenomena that affect everything from household water flow to the combustion processes in jet engines.
Turbulence, characterized by chaotic fluctuations in time and three-dimensional space, remains one of the most challenging aspects of fluid dynamics. The complexity inherent in turbulence has long puzzled scientists, despite its importance in numerous practical applications.
According to P.K. Yeung, a professor in the Daniel Guggenheim School of Aerospace Engineering, with a courtesy appointment in the George W. Woodruff School of Mechanical Engineering, the breakthrough achieved on Frontier could pave the way for new discoveries.
“Turbulence is very complex, theories are incomplete, and laboratory measurements are arduous,” Yeung said in a news release. “A world-leading resolution of over 35 trillion grid points on Frontier is expected to lead to new discoveries, which in turn can facilitate advances in modeling where both assumptions and predictions can be tested numerically.”
The achievement was made possible through the capabilities of Frontier, located at the Oak Ridge National Laboratory. This supercomputer, which can perform a quintillion operations per second, enabled Yeung and his team to execute simulations that were previously thought unimaginable. Frontier’s powerful graphical processing units (GPUs) play a critical role in these high-resolution computations.
Yeung’s team accessed Frontier at its inception and benefited from significant time allocations through the INCITE (Innovative and Novel Computational Impact on Theory and Experiment) program, managed by the U.S. Department of Energy’s Office of Science.
Their pioneering work was detailed in a recently published journal article, outlining a specialized algorithm designed to maximize Frontier’s computational features for highly efficient, extreme-scale simulations.
“In many scientific fields, people thought calculations of this magnitude were not possible, but now we are there, perhaps earlier than anticipated,” Yeung added. “Our work on turbulence simulations also demonstrates several principles of advanced GPU programming of interest in other fields, especially those where so-called pseudo-spectral methods are important.”
Beyond the immediate scientific insights, the data generated from these simulations will be shared publicly through a collaboration with the National Science Foundation-supported Johns Hopkins Turbulence Database project, further enriching the global scientific community’s resources.
The advancements achieved by Yeung and his team at Georgia Tech not only mark a milestone in turbulence research but also highlight the transformative potential of next-generation computational resources in broadening our understanding of complex physical systems.