Scientists and researchers are racing against the clock to find a cure for COVID-19. The search requires the screening of billions of drug candidates and identifying the right chemical structure that will most favorably bind—and interfere—with the virus.
However, calculating the binding potential is computationally intensive, taking years to screen a billion compounds on a CPU system. The GPU accelerates this process.
This demo shows that finding a drug for COVID-19 is challenging because the pool of potential drugs is enormous and calculations needed to identify the odds of a drug binding to the virus are compute intensive. However, GPUs accelerate this process, allowing researchers to screen over 1B candidates in 12 hours on the Oak Ridge National Lab’s Summit supercomputer.
Over 700 applications are ported to GPUs, giving researchers in various scientific domains the tools needed to accelerate their workloads and make groundbreaking discoveries faster.