Researchers from the Massachusetts Institute of Technology (MIT) developed an artificial intelligence approach to detect electron correlation, which could help further computational chemistry.
MIT's Heather Kulik and colleagues used the Comet supercomputer at the University of California, San Diego's San Diego Supercomputer Center and the Bridges supercomputer at the Pittsburgh Supercomputing Center in this effort. The resulting artificial neural network models predict strong correlation in materials at significantly lower computational cost than conventional models, potentially accelerating the search for materials in diverse applications.
The MIT team's MultirefPredict workflow engaged with at least three electronic structure codes and utilized central processing units and graphics processing units on Comet and Bridges. "Using those supercomputers firsthand allowed me to think about ways I can teach students who may just be learning computational chemistry to complement their experimental research for ways that they can use not only now but in the future," Kulik says.
From University of California San Diego
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA