My office is littered with notebooks, mainly Moleskines as I’m a colossal hipster nerd (sans beard) and I only write in them with Uniball Eye Micro pens (I knew you were all thinking it).
On the 6th June 2013 I scribbled some notes about using mobile phones as cluster node devices to do small chunks of processing, I even wrote a blog post. My rationale was simple, with all the phones on the planet in standby wouldn’t it make sense to use the downtime like we did with screen savers, trying to solve a bigger problem, a medical one for instance.
It didn’t get very far in my mind, too complex and while a few had tried Hadoop on Android the phones just couldn’t handle the load in a decent way. I toyed with using Zookeeper or RabbitMQ to do the communication work. Regardless of which way I was to do it, it was going to be hard for one man and a kitchen table to do this sort of thing. Not impossible, just hard.
Mistake 1 – I kept the idea to myself.
Well that’s not 100% true, I did email one person and tell another. On the grounds of it being too complex for one person to handle I closed the notebook and left it at that, a fresh page waiting for the next idea.
That was a mistake….
Mistake 2 – Didn’t share with my collaboration partners
When I say “collaboration parters” I mean, my developer network, me mates. Perhaps they’d see something I didn’t. Was I just way ahead of the right time (usually the way, far too early stage for my own good sometimes). Regardless I didn’t show it to them.
Mistake 3 – Believing your own internal critic
After staring at the said page in my notebook for a few days I left it as a bad idea. Too ahead of the curve, limited use and I couldn’t see that anyone would be interested.
Sometimes the inner rascal will tell you something is right when it’s wrong, sometimes it will tell you the opposite.
With all that in mind, I just found out something.
Turns out I was completely wrong……
The new issue of Wired landed on my door mat this afternoon, yes the print edition, can’t be doing with digital edition of these sorts of things (apart from HBR). At the bottom of one page something about an app that cures cancer while you sleep. The premise is built on using mobiles as a cluster….. beautiful, and the numbers are highly encouraging. “…at one point 53,000 phones were working at one time. This is twice as fast as any supercomputer in the US“.
So Folding@Home out of Stanford University. Well done for daring, poking the devices and seeing what was possible. Vijay Pande I salute you! Well done and brilliant work.
If you want to download the app and putting your device to work, have a look at the Folding@Home website.