How the pandemic revived a distributed computing project and made history [View all]

[LiberalArkie]

Almost 20 years ago, faculty in the chemistry department of Stanford University launched a distributed computing project called Folding@Home (F@H). They sought to understand how proteins self-organize and find out why this process sometimes goes wrong, causing issues such as cancer and Alzheimer’s Disease.

F@H hit its pinnacle of mindshare—and performance—in 2007, when Sony added it to the PlayStation 3. But like many other projects, it saw a gradual decline in its popularity since. This past March, however, F@H saw a sudden resurgence. Thanks to a confluence of events, notably including the SARS-CoV-2 pandemic, Folding@Home broke the exaFLOP barrier at least one or two years before Intel, AMD, IBM, or Cray could do it. Here’s how those events played out.

Snip

Then in February, everything changed. Folding@Home suddenly went from 30,000 volunteers running the software in February to 400,000 in March—another 300,000 users came on board after that. There were so many users that the database ran out of potential simulations for them to crunch, and data coming in was so great that the servers were overloaded, said Bowman.

Despite these glitches, F@H zoomed to a peak performance of 1.5 exaFLOPs, making it more than seven times faster than the world's fastest supercomputer, Summit, at the Oak Ridge National Laboratory.



More

https://arstechnica.com/science/2020/04/how-the-pandemic-revived-a-distributed-computing-project-and-made-history/