The Pande Lab at Stanford University has been running the [email protected] project for almost twenty years. The distributed compute network utilizes volunteers’ spare CPU and GPU power, combining the resources of thousands of home systems to understand better how proteins fold. Hopefully, the simulations can lead to treatments for diseases such as cancer and Parkinson’s.
Last month saw [email protected] thrust into the spotlight after it added Covid-19 to the list of diseases it’s researching. In less than a week, 400,000 people had signed up to lend their system resources to the project, giving it access to 470 PetaFLOPS of power. Another week later, the cumulative performance reached 1.5 ExaFLOPS, or 1,500,000,000,000,000,000 floating point operations per second.
Now, [email protected] has passed an incredible 2.4 ExaFLOPS. IBM’s Summit, which tops the most powerful supercomputers list, has a peak output of 200 PetaFLOPS and a LINPACK benchmark of 148.6 PetaFLOPS.
[email protected]’s director, Greg Bowman, tweeted that the service has had to refocus its efforts from setting up new projects to moving data off servers to make room for more. With around 6TB of data arriving every hour, it’s not surprising that the project needs more space.
If you want to become part of [email protected], just download the client. You’ll be joining several tech outlets and hardware vendors, including Nvidia and EVGA, that have put teams together.