kentoh - Fotolia
In the race to flatten the curve of COVID-19, high-profile tech companies are banking on supercomputers. IBM has teamed up with other firms, universities and federal agencies to launch the COVID-19 High Performance Computing Consortium.
This consortium has brought together massive computing power in order to assist researchers working on COVID-19 treatments and potential cures. In total, the 16 systems in the consortium will offer researchers over 330 petaflops, 775,000 CPU cores and 34,000 GPUs and counting.
COVID-19 High Performance Computing Consortium
The consortium aims to give supercomputer access to scientists, medical researchers and government agencies working on the coronavirus crisis. IBM said its powerful Summit supercomputer has already helped researchers at the Oak Ridge National Laboratory and the University of Tennessee screen 8,000 compounds to find those most likely to bind to the main "spike" protein of the coronavirus, rendering it unable to infect host cells.
"They were able to recommend the 77 promising small-molecule drug compounds that could now be experimentally tested," Dario Gil, director of IBM Research, said in a post. "This is the power of accelerating discovery through computation."
In conjunction with IBM, the White House Office of Science and Technology Policy, the U.S. Department of Energy, the National Science Foundation, NASA, nearly a dozen universities, and several other tech companies and laboratories are all involved.
The work of the consortium offers an unprecedented back end of supercomputer performance that researchers can leverage while using AI to parse through massive databases to get at the precise information they're after, Tim Bajarin, analyst and president of Creative Strategies, said.
Supercomputing powered by sharing big databases
Bajarin said that the world of research is fundamentally done in pockets which creates a lot of insulated, personalized and proprietary big databases.
"It will take incredible cooperation for Big Pharma to share their research data with other companies in an effort to create a cure or a vaccine," Bajarin added.
Gil said IBM is working with consortium partners to evaluate proposals from researchers around the world and will provide access to supercomputing capacity for the projects that can have the most immediate impact.
Many enterprises are coming together to share big data and individual databases with researchers.
Signals Analytics released a COVID-19 Playbook that offers access to critical market intelligence and trends surrounding potential treatments for COVID-19. The COVID-19 Playbook is available at no cost to researchers looking to monitor vaccines that are in development for the disease and other strains of coronavirus, monitor drugs that are being tested for COVID-19 and as a tool to assess which drugs are being repurposed to help people infected with the virus.
"We've added a very specific COVID-19 offering so researchers don't have to build their own taxonomy or data sources and can use it off the shelf," said Frances Zelazny, chief marketing officer at Signals Analytics.
Eschewing raw computing power for targeted, critical insights
With the rapid spread of the virus and the death count rising, treatment options can't come soon enough. Raw compute power is important, but perhaps equally as crucial is being able to know what to ask and quickly analyze results.
"AI can be a valuable tool for analyzing the behavior and spread of the coronavirus, as well as current research projects and papers that might provide insights into how best to battle COVID-19," Charles King, president of the Pund-IT analysis firm, said.
The COVID-19 consortium includes research requiring complex calculations in epidemiology and bioinformatics. While the high computing power allows for rapid model testing and large data processing, the predictive analytics have to be proactively applied to health IT.
Dealing with COVID-19 is about predicting for the immediate, imminent future -- from beds necessary in ICUs to social distancing timelines. In the long term, Bajarin would like to see analytic and predictive AI used as soon as possible to head off future pandemics.
"We've known about this for quite a while -- COVID-19 is a mutation of SARS. Proper trend analysis of medical results going forward could help head off the next great pandemic," Bajarin said.