Putting HPC To Work To Accelerate COVID-19 Research

by | Apr 23, 2020 | AI and Machine Learning, In the News

The US arsenal of supercomputers has officially been opened up for COVID-19 research. The White House recently announced the new COVID-19 High Performance Computing Consortium, which will allow researchers worldwide to access to the world’s most powerful HPC resources and leverage them to combat the novel coronavirus.

The White House believes that these high performance computing resources can significantly advance the pace of scientific discovery in the fight to stop the virus. The consortium includes government researchers from the five DOE research labs, home to the largest supercomputers in the word, as well as companies like NVIDIA, IBM, Microsoft, Amazon Web Services, Google and HPE. Scientists from MIT, RPI, NASA and the National Science Foundation help round out the extensive team.

Figure 1: The Summit Supercomputer at the Oakridge National Lab facility in Tennessee is the fastest supercomputer in the world. image: NVIDIA

“The COVID-19 HPC Consortium is the Apollo Program of our time,” Ian Buck, NVIDIA VP and GM for Accelerated Computing, said. “Not a race to the moon, this is a race for humanity. The rocket ships are GPU supercomputers, and their fuel is scientific knowledge. NVIDIA is going to help by making these rockets travel as fast as they can.”

To help the global team tackle the virus, NVIDIA is providing GPU and AI expertise, free software such as the Parabricks genome analysis toolkit, and optimization assistance for computing at massive scale. This help can be a game changer. A 20% improvement in scalability on a 330 Petaflop supercomputer can essentially free up another supercomputer’s worth of power, equivalent to the fourth largest supercomputer in the world.

Using supercomputers to aid in drug discovery and efficacy simulation could dramatically accelerate the time it takes to find and validate a solution to the global pandemic that has smashed the world’s economy and brought so much misery to people’s lives. Using software like Parabrix on GPU-accelerated computers can speed up the process by over 50X. A research team using the world’s fastest supercomputer, Summit, at Oak Ridge National Laboratory, has already identified 77 drug compounds that could be effective for treating the disease. Powered by more than 27,000 NVIDIA GPUs and over 9000 IBM POWER CPUs, Summit helped reduce discovery time from years to a few days. Additionally, using AI to scan through millions of potential drugs can shave months off the time needed to evaluate the potential universe of medications.

Wrapping up

As we’ve all heard so many times, we are all in this together. Let’s give our scientists and medical professionals the very best tools available to help them stop this disease before even more damage is done. Thanks to everyone who is contributing to these efforts. Respect.