Page 1 of 1

Re: Recent computer advances. [URLs]

Posted: Wed Apr 17, 2013 4:26 pm
by Jesse_V
Adam A. Wanderer wrote:The F@H program needs to be one of ever increasing power and flexibility.
Folding@home: Lessons From Eight Years of Volunteer Distributed Computing (2009)
The main challenge of large local computation is the cost for the hardware, electricity, cooling, floor space, and system administrators. By using distributed computing, we can eliminate many of these costs completely because we use existing resources. The cost of the volunteer's computer and Internet connection is already spent since they need a computer for other reasons. This makes the only additional cost the difference in electrical use of the busy system versus idle ones. This per-volunteer cost has been considered insignificant by those participating, but may become a concern if electrical costs increase greatly in a geographic area.

The management of clients by volunteers gives us the advantage of a self-upgrading system. As volunteers upgrade their systems, Folding@home is upgraded in capacity as well. As a whole this gives us Moore's law growth without any additional effort. As new architectures are released such as the recent move form PPC to x86 by Apple or the move from 32 to 64bit Linux, we have to do the work of porting and testing the cores and recompiling the client, but the computers are upgraded over time without our intervention.

Re: Recent computer advances. [URLs]

Posted: Wed Apr 17, 2013 11:22 pm
by bruce
I'm not in a position to be a fanboi for or against Intel or GPU technology as certain to turn out to be the better technology. [Slashdot is, which is why they say "Sponsored by Intel" on their website.]

Recognize it for what it is: Salesmen promoting a new technology which may or may not "win" in the marketplace compared to some other technology in the marketplace that's just as likely to "win" in some sense. They're not wrong, but they're equally not right. Nobody can accurately predict.