Folding@Home and Rosetta@Home projects aim to perform biomedical research using the computing power of volunteers. Basically, you just need to install a program on your computer, and it will use idle computing power to perform complex calculations without slowing down your computer as long as you are not short in RAM.
The projects are now working on COVID-19 to understand how SARS-CoV-2 protein is structured which could help find a cure. The programs have been available for Windows, Linux and Mac OS on 32-bit and 64-bit x86 targets for years, but very recently Rosetta@Home has been made available for 64-bit ARM targets so people can also run BOINC program on Arm Linux SBCs such as Raspberry Pi 4, NVIDIA Jetson Nano, or Rock64, or even powerful Arm servers to help with Rosetta@Home project’s COVID-19 research.
As explained in an article on miniNodes, you’ll need a board with at least 2GB RAM and running a 64-bit operating system. That means Raspbian will not work since it’s only 32-bit, and instead you can use Ubuntu 18.04 / 19.04 server 64-bit for Raspberry Pi. On other SBC’s, people have been using Armbian successfully.
On Debian/Ubuntu-based operating systems, you can install BOINC as follows:
sudo apt-get install boinc-client boinctui
Not that BOINC for 64-bit Arm has been around for several years, but Rosetta@Home has only recently started to send work units to 64-bit Arm devices.
You can then use your computer to go to create a BOINC account and join a team such as “crunch-on-arm” team. Once it’s done, you can run BOINC Terminal UI program in your Arm device:
Then press F9 to access the menu, navigate to Projects, select Add Project, and select Rosetta with Existing User and enter the credentials you just created in the previous step.
After a while, Rosetta should begin downloading the necessary files, then some work units, before crunching data on your board. You can press ‘Q’ to exit the UI, and it will still run in the background.
Note that if you are using a Raspberry Pi 4 with 2GB RAM only, only 1.9GB RAM may be detected, and you may have to tweak your config files. This is explained in miniNodes post. That also means some other SBC’s with 2GB RAM may have issues, so using boards with 4GB RAM or more if a safer bet.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.