LEMONT, Ill. — Instead of radar, scientists at Argonne National Laboratory are using a supercomputer. And it makes some spectacular and speedy calculations, including the prediction that COVID-19 will not go away.
THETA is one of the fastest, most powerful supercomputers in the world. It hums along at the Argonne lab in suburban Lemont and spits out massive data sets the brilliant minds interpret.
Charles Macal, PhD is a senior system engineer at Argonne.
“The computer doesn’t have consciousness. It doesn’t tell us what to do, but it’s very helpful,” he said. “We’re doing things now that no one else in the world is doing.”
They’re like weather forecasters. Except the storm Macal and his team are tracking is Covid.
“Something early on that really knocked my socks off was the realization that the spread of COVID-19 the virus will simply go on and on and on,” he said. “That’s something I never expected. You thought it would die out. You hope it’s going to die out.”
They started modeling last March and fed THEATA’s insatiable Appetit for information. What takes the supercomputer 50 to 60 hours to calculate, Macal said would take a computer simulation much longer.
“We estimate that to do this same work, a computer simulation of COVID-19 on a laptop would take 20 years to turn out those results,” he said.
PPE, ICU beds and even body bags factored into the early equations. The team shared the projections with decision makers at local and state health departments.
“Our first finding was wearing masks and doing social distancing could reduce transmission from person to person by 85 or 90 percent,” Macal said. “That was our first very striking and very strong and clear signal of the numbers coming out of our model and forecasts.”
Over the summer, they used the supercomputer to model the impact of protests.
“We found that there would be very little impact of the protests on the transmission,” Macal said.
And they ran different scenarios related to schools reopening. Their prediction with the proper protective behaviors in place?
“Schools would not be a major source of spreading the disease,” Macal said. “And now we have the advantage of seeing how the data unfolded. And the experiences were apparently, from the data, very good in terms of limited or very few outbreaks.”
Now the Argonne team is focused on SARS-CoV-2 mutations. The latest models from theta make a bold prediction when it comes to the virus variants in circulation right now.
“The variant in the UK led to 70 percent more people going to the hospital,” Macall said. “So we ran that through our model in the past week. The variant is more transmissible, so it will kind of win out in the race to infect people versus what we call the original strain. It could cause another wave, in particular of hospitalizations, because the new infections have a higher proportion of people going to the hospital. The projections are showing that this could happen not that it will happen, but that it’s in the realm of possibility.”
The Argonne team is also working on projecting the impact of the vaccination program, taking into account the uncertainties of the supply chain and the impact of variants.