RIP [email protected] I Didn’t Really Understand You, But I’ll Miss You

RIP Folding@Home. I Didn’t Really Understand You, But I’ll Miss You

Remember [email protected]? AKA Life With PlayStation? If not, it would be hard to blame you. It was something I downloaded but never really took the time to understand. Now Sony has announced that, in the beginning of November, it will be retiring [email protected] from the PlayStation 3. You will no longer be able to help Stanford University understand protein folding and further Alzheimer’s research by casually opening an application on your PlayStation 3.

It was an interesting system. Post the 1.6 firmware update, which was made available all the way back in 2007, the [email protected] icon was added to the PlayStation’s Cross Media Bar. All folks needed to do was select it, and watch as the PlayStation 3 began calculating ‘stuff’. The idea was that, with multiple machines linked together, working in tandem, previously unmanageable calculation would become routine, helping Stanford University in its research.

It was something I never really properly understood, not really, but it was enough to know that I was kinda involved, somehow. It was a weird kind of interactivity, watching other PlayStation users connect across the world.

It’ll be a shame to see this strange thing go.

PS3 System Software Update (v4.30) [PlayStation Blog]


  • You can still help with your PC if you want to 😉 There is plenty of projects needing processors 🙂

    BONIC and BAM! 🙂

  • The idea I believe is that there’s nearly infinite possibilities for configuring a protein but only some of them that will fold correctly. Evaluating them one by one on a single computer would probably not finish for billions of years. However if you could do two at once, you would halve the time taken to find the ‘good’ configurations, and the more configurations you can do simultaneously, the faster you find the good ones. So if you had a billion computers to do a task that sequentially takes a billion years, you would complete it in one year. Search problems like this often can be done in parallel like this really effectively.

    • That’s basically right. If a million PS3s were doing these calculations in theory it’d take one millionth of the time to find the relevant information.

    • Pretty much spot on.

      Really sad to see this go, it was a great idea, and it really did contribute to something we could all benefit from. imagine what progress we could make if all consoles took part on all different systems, and all it took was to turn on your console and nothing more.

      There’s another similar program which uses the same principle for Astronomy. Funnily enough it’s also called “theSkynet” (if mass connected computers already didn’t worry you :P)

  • My question is, why does it use GPU’s? It seems like a task best for a CPU surely?

    I’ve had [email protected] explained to me numerous times, but I just don’t understand it. People compete in teams, but for what exactly?

    • The basic explanation is, well, you know how a CPU can have 2 or 4 cores. A modern GPU has thousands. They’re much smaller, though, and can only do some specialised mathematical calculations. A GPU core is not good for running programs, but is great for doing graphical calculations. Or, as it turns out, protein folding calculations as well.

    • Absolute speculation (absolutely absolute), but I think just maybe GPUs are better at things like this because they can do things simultaneously?

  • Oh man that sucks 🙁

    I actually referred to it as Playstation Heater during winter months! Would always keep my room toasty on a cold night :p

    In all seriousness though, I’m over 1000 work units completed on my ps3. I had the standby timer which would autostart it if I didn’t touch my ps3 in 10 minutes, so it got quite a lot of use.

    I still run BOINC on quite a few PC’s, which I recommend everyone else to do also, but was always good knowing my CELL processor was always being used to it’s full potential.

    It is also vague on the reasoning for discontinuing the program. While first reflex would be to blame Sony, after a bit of thought I imagine that the guys at Stanford probably don’t want to have to worry about supporting the platform still. Most distributed computing projects are all starting to focus on GPU utilisation, which is many orders of magnitude more efficient than even the PS3’s CELL.

    Farewell old friend!

    • I believe it just used the SPE’s for the bulk of the processing and the PPE for process control [much like most ps3 games]. As far as I know the RSX does not have support for any GPGPU application, as such it was just used for visualisation only.

Show more comments

Log in to comment on this story!