Skip to main content
This protein drives key processes for sulfide use in many microorganisms that produce methane, including Thermosipho melanesiensis. Researchers used supercomputing and deep learning tools to predict its structure, which has eluded experimental methods such as crystallography.  Credit: Ada Sedova/ORNL, U.S. Dept. of Energy

A team of scientists led by the Department of Energy’s Oak Ridge National Laboratory and the Georgia Institute of Technology is using supercomputing and revolutionary deep learning tools to predict the structures and roles of thousands of proteins with unknown functions.

Using quantum Monte Carlo methods, the researchers simulated bulk VO2. Yellow and turquoise represent changes in electron density between the excited and ground states of a compound composed of oxygen, in red, and vanadium, in blue, which allowed them to evaluate how an oxygen vacancy, in white, can alter the compound’s properties. Credit: Panchapakesan Ganesh/ORNL, U.S. Dept. of Energy

Neuromorphic devices — which emulate the decision-making processes of the human brain — show great promise for solving pressing scientific problems, but building physical systems to realize this potential presents researchers with a significant

ORNL’s Steven Young (left) and Travis Johnston used Titan to prove the design and training of deep learning networks could be greatly accelerated with a capable computing system.

A team of researchers from the Department of Energy’s Oak Ridge National Laboratory has married artificial intelligence and high-performance computing to achieve a peak speed of 20 petaflops in the generation and training of deep learning networks on the