Folding Farm

For a while I got quite interested in the Folding@Home project, where you use home computers to figure out the way certain proteins are folded in order to aid medical research. It’s a way for medical researchers to utilise all the untapped computing power in people’s desktop computers all over the world.

I took it a step further by collecting old computers and networking them in to a mini-datacentre in my shed (though “datacentre” is overselling it a bit). I tried to install a different operating system on each one (different flavours of Linux, FreeBSD and Windows) so that I would have a bit of experience with each one and better understand the different environments. I ended up with 22 hosts in total, busily folding away at proteins. It became a bit of a challenge to see if I could pick up some decent hardware for less than $50 through the used-PC classifieds.

I wrote some monitoring software to keep track of all the machines: which protein they were working on, their current status and how long they’d been running. It was quite satisfying to think of all that work being done 24/7 in my shed.

In the end I decommissioned the farm and chucked out all the machines for a few reasons:

  • standard computing power crept forward so that my farm was really behind the curve – I was no longer getting enough computing power for the running costs relative to what everyone was able to achieve (my contribution was getting relatively smaller and smaller each day)
  • I wanted to use my shed for other things!

I encourage you to use your computers to do some folding though, you don’t need a big cluster to be useful. You can use your computer’s spare cycles tohelp find cures for diseases like Parkinson’s and Alzheimer’s.

The shelves full of servers in my shed

The shelves full of servers in my shed

A screenshot of the monitoring software

A screenshot of the monitoring software (click to embiggen)

Leave a Reply