An application that every inhabitant of the planet can install on their computer has helped build a network whose computing power has recently exceeded the total capacity of the five hundred most powerful supercomputers in the world.
20 years ago, researchers at Stanford University developed the Folding @ Home application, which any PC user can install on their computer. When the performance of the machine is not used, as during sleep or reading paper books, or is not needed completely, as when watching a series in low quality, the user can provide his machine for medical research. All you need is the Internet and a built-in computer schedule.
According to director of Folding @ home biochemist Greg Bowman, the COVID-19 pandemic has dramatically increased the number of people who want to share the power of personal computers with a research center. The number of “donors” of capacities increased from 30 to 700 thousand people. Earlier this week, the computing power of computers connected by an application around the world exceeded 2.4 exaflops.
Flops is an off-system unit showing the number of computational operations per second. The prefix “exa” means that every second, voluntary computer support for the inhabitants of the globe provides physicians with a quintillion or million million million operations per second. The peak power of the world’s most productive supercomputer, which occupies the top position in the top 500 ranking, is “only” a couple of hundred teraflops (“tera” is a trillion or a million million). The total performance of all five hundred cars on the list – 1.65 exaflops – is less than that collected from the world by thread.
The Folding @ home application breaks down a complex scientific task into small subtasks and distributes them among users’ accessible computers. Actual at the moment is modeling the behavior of viruses, for example, SARS-CoV-2, in living organisms. Scientists are interested in the details of the penetration of three coronavirus proteins into human cells. Understanding this process significantly accelerates the release of the vaccine, the effect of which on the body can also be modeled using a supercomputer distributed around the world.
Researchers hope to soon increase the number of personal computers involved in the calculations to one million.