Even the toughest materials on Earth are vulnerable under the most extreme conditions. A supercomputer simulation visualized just that by cracking, melting and recrystallizing a virtual diamond under immense pressure and unimaginable temperatures.
“We created gigantic simulations of a micron-sized hunk of compressed diamond,” explained Aidan Thompson, a scientist at Sandia and a member of the research team, which was led by University of South Florida (USF) professor Ivan Oleynik. “To do this, we track the motion of billions of atoms by repeatedly calculating the atomic forces over very many, exceedingly tiny, intervals of time.”
Thompson is the originator of SNAP (short for Spectral Neighbor Analysis Potential), a model that Sandia uses to quickly predict how billions of atoms — like those in this research — interact with one another. The simulations were run on Summit, which remains the most powerful supercomputer in the United States at 148.6 Linpack petaflops, over the course of a day. They show the microscopic chunk of diamond, under 12 megabars of atmospheric pressure and at 8540° Fahrenheit, cracking, melting into shapeless carbon and recrystallizing.
But even working on a powerhouse like Summit, simulations like this one need clever optimization. “Since 2018, just by improving the software, we have been able to make the SNAP code over 30 times faster, shortening the time for these kinds of simulations by 97 percent,” Thompson said. “At the same time, each generation of hardware is more powerful than the last. As a result, calculations that might have until recently taken an entire year can now be run in a day on Summit.”
For this simulation in particular, the Sandia team began with quantum mechanical calculations that are only feasible for some hundreds of atoms. A separate model was then trained using machine learning to scale the initial model up to a system containing billions of atoms. “Since supercomputer time is expensive and highly competitive,” Thompson said, “each shortening of SNAP’s run time saves money and increases the usefulness of the model.”
And, strange as it may seem, there are plenty of applications for this sort of research.
“We can now study the response of many materials under the same extreme pressures,” Thompson said. “Applications include planetary science questions — for example, what kind of impact stress would have led to the formation of our moon. It also opens the door to design and manufacture of novel materials at extreme conditions.”
The research team also included members from the National Energy Research Scientific Computing Center (NERSC) and Nvidia. The research was selected as a finalist for the Gordon Bell Prize in 2021. To learn more about the research, read the technical paper here or read more at this link.
Header image: uncompressed diamond (blue), cracks (red) and the final state (orange). Image courtesy of Sandia National Laboratories.