Augustine wrote:

that was a typo in the url. remove the final ‘l’, you’ll get the article.

Yeah, fixed.

Wow. Didn’t realize smartphone processors were that powerful.

I recall that:

iPhone 4: 800MHz
iPad: 1GHz

The more important question is if the problem being solved can be subdivided adequately so that it can run on multiple devices with little or more likely no interaction between devices. Things like SETI@home could work, complex simulations probably won’t.
The article doesn’t make clear to me whether the bulk of the processing is happening in the device, or whether it’s shared. And the emphasis seems to be on the software iterations, which progressively trim the error. The “smartphone” theme or example may just have been to emphasize the power of the software; it may work much better on more powerful computers.

Which is why I thought it might have applicability to LPP’s simulations.

Henning wrote: I don’t know about this particular article, but I read an article recently where the mobile phone is just the remote control of a supercomputer. Wouldn’t make much sense for our kind of simulations anyway, because you’ll need tight coupling of the processors.

I think the “mobile phone” stuff is just an eye-catching red herring. The real point is the software, which uses advanced new algorithms.