NVIDIA's newest AI supercomputer joins the fight against COVID-19

The DGX A100 offers 5 petaflops of processing power.

Radoslav Zilinsky via Getty Images

NVIDIA is bringing its AI expertise to bear in the fight against COVID-19, the company announced during its GTC 2020 keynote on Thursday. Specifically, the first of NVIDIA’s new line of AI-driven supercomputing systems, the DGX A100, will be sent to Argonne National Lab where it will help government researchers screen potential coronavirus therapeutics far faster than they could do by hand.

The DGX A100 is NVIDIA’s third generation AI supercomputer. It boasts 5 petaflops of computing power delivered by eight of the company’s new Ampere A100 Tensor Core GPUs. A single A100 can consolidate “the power and capabilities of an entire data center into a single flexible platform for the first time,” an NVIDIA press release claimed.

“We’re using America’s most powerful supercomputers in the fight against COVID-19, running AI models and simulations on the latest technology available, like the NVIDIA DGX A100,” Rick Stevens, associate laboratory director for Computing, Environment and Life Sciences at Argonne, said in that statement. “The compute power of the new DGX A100 systems coming to Argonne will help researchers explore treatments and vaccines and study the spread of the virus, enabling scientists to do years’ worth of AI-accelerated work in months or days.”

Once the A100 is installed in Argonne, researchers there will “be able to screen 1 billion drugs in under 24 hours,” Kimberly Powell, Vice President of Healthcare at NVIDIA, told reporters in a briefing. Without the added computing power, researchers would need close to a year to screen the same number of therapeutics.

“This supercomputer is going to combine both accelerated computing and artificial intelligence,” Powell continued. “Critical applications in molecular dynamics simulation, where you can understand how molecules interact with drugs, how to interfere with the virus protein binds to the cell protein and how to block it with a molecule. And then how can we screen as many drugs as possible so we can more quickly get that drug candidate into experimentation and into clinical trials.” Up to 140 A100s can be run in parallel, pooling the combined power of their GPUs into a “DGX SuperPOD” that boasts more than 700 petaflops of processing capability.

NVIDIA also announced a host of other new services it plans to roll out to hospitals across the country. Using its line of Clara Parabrick GPUs, NVIDIA has managed to reduce the time needed to analyze an entire human genome from around 30 hours to less than 20 minutes. The company is also teaming with more than a dozen industry partners and 50 hospitals around the globe to help better protect frontline medical staff through the use of remote monitoring, video conferencing and at-a-distance body temperature monitoring in a program called NVIDIA Clara Guardian.