Would you loan your PS3 to scientists?

Started by
24 comments, last by BleedingBlue 17 years, 7 months ago
Quote:Original post by Yann L
Anyway, let's assume for a moment their claims were real and not marketing blubber. So, if the PS3 architecture was that powerful, why all this huge effort for internet based distributed computing ? Why not just build a massive parallel computer from those PS3 cell processors ?

:)
your wish is granted

Quote:Computer giant IBM will build the world's most powerful supercomputer at a US government laboratory.

The machine, codenamed Roadrunner, could be four times more potent than the current fastest machine, BlueGene/L, also built by IBM.

The new computer is a "hybrid" design, using both conventional supercomputer processors and the new "cell" chip designed for Sony's PlayStation 3


http://news.bbc.co.uk/2/hi/technology/5322704.stm

Advertisement
Quote:Original post by MichalsonNoted computer scientist David Manning from Ridgefield University has said that the Playstation 3 is the most powerful computer he's ever seen.


LOL!!! I had to search the name to get it, but that is hilarious!
Out of curiosity, how does information processing work in terms of energy efficiency?

True, there is tons of waste heat, but if you were using a computer as a replacement for a heater and all that heat was the desired outcome, is the processing that gets done in the process essentially free, and if not, what did the heat we should have gotten instead become?
Chess is played by three people. Two people play the game; the third provides moral support for the pawns. The object of the game is to kill your opponent by flinging captured pieces at his head. Since the only piece that can be killed is a pawn, the two armies agree to meet in a pawn-infested area (or even a pawn shop) and kill as many pawns as possible in the crossfire. If the game goes on for an hour, one player may legally attempt to gouge out the other player's eyes with his King.
Quote:Original post by smart_idiot
Out of curiosity, how does information processing work in terms of energy efficiency?

True, there is tons of waste heat, but if you were using a computer as a replacement for a heater and all that heat was the desired outcome, is the processing that gets done in the process essentially free, and if not, what did the heat we should have gotten instead become?


Well, if you have blank hard drive, and write results of computation to it, some very very small energy comes into magnetic field. Rest just becomes heat, thanks to conservation of energy.
That's said, if a bit of information is irreversibly 'destroyed', some energy (log(2)*k*T or something like that) must be released as heat. The computation in which no bits are irreversibly lost may theoretically run absolutely 'for free' not even releasing any heat.

Also, it is in theory possible (but not always practically viable) to either a: heat your room directly without inefficiency of steam turbines or other means of electrical generation (more efficient than having power plant that powers your computer). or b: use heat pump that are more efficient than just heater (most heat pumps won't work when it's -30c outside tho).
Quote:Original post by Yann L

Essentially, they are saying that this PS3 setup, at a price of around $13m, would outperform all supercomputer resources in the US. Mind you, we're talking about hundreds of millions, if not billions of Dollars here.

Yeah, riiight...


That's not what the article said. The scientists said that with 10000 PS3's, THEIR network would perform five times faster, which is a very reasonable claim ...
h20, member of WFG 0 A.D.
I think no matter what distributed computing over the internet will always be cheaper then giant local clusters. For a few reasons.

1) I and the many other contributors to the project power our own respective node of the cluster. This means of course that I bear the cost of powering my node, not the project.

2) By having many nodes distributed over the world they (seti@home et al) do not have to provide cooling, further cutting their own energy costs down.

3) The cost to add additional nodes to their systems is marginal (mostly bandwidth costs), potentially additional servers to handle additional load.

So even if they could theoretically build a massive super computer for a marginal (marginal being relative to the cost of larger scale supercomputers), a distributed model still has many benefits over a large local cluster.

That and even if it were a bunch of marketing BS, I'd still contribute my hardware to the cause, at the very worst, my ps3 wastes some clock cycles contributing to a worthwile cause, which isn't to disimilar from not contributing, where my ps3 wastes some clock cycles contributing to some not so worthy causes, like entertaining me.

This topic is closed to new replies.

Advertisement