Distributed Computing made EZ

[dvandom]

If any of you have some older machines sitting around and would like to try making a simple parallel processor out of them, check out the DCEZ project at Emporia State. It only works for "embarrasingly parallel" problems, but that includes POV-Ray rendering. You can set up a parallel network in 5-10 minutes with this, according to the talk I was at last weekend. (I meant to mention it earlier, but forgot...just got that far in typing up my notes to be reminded today.)

Comments

[jarodrussell.livejournal.com]

Would that project help physicists any?

[dvandom.livejournal.com]

Yep. A lot of computational physics involves what are called Monte Carlo simulations, which take up loads of processing power, but can help with experiments at levels from upper division undergrad to cutting edge research. Heck, just sticking this on the machines in the physics student computer lab and running cycles when there isn't a class in there would provide enough power for serious research.

[jarodrussell.livejournal.com]

*nods* *nods*

Cool. I knew grid computing (http://en.wikipedia.org/wiki/Grid_computing) was helpful to geneticists and astronomers, and I knew that places like CERN had some serious computing power, but I just never really hit me that grad and post-grad physicists could make use of it...

That, and I wasn't sure if that specific package would be very useful.

THANK YOU! That's very interesting to know.