[ExI] Life @ Playstation

[Tomasz Rola]

On Sun, 4 Nov 2012, Stefano Vaj wrote:

> On 4 November 2012 04:07, Tomasz Rola <rtomek at ceti.pl> wrote:
> > My very wild guess, PS3 is no longer significant for this kind of task.
> 
> 
> Mmhhh.
> 
> It would appear that as of yesterday 172244 Windows CPUs deliver 179
> Teraflops to the project, 16557 PS3s deliver 985 x86-equivalent
> Teraflops, which makes for a couple of orders of magnitude greater
> performance for PS3
> 
> Morever, the project appear to have still some 2000 PowerPC Macintosh
> (!) contributing a couple of Teraflops, even though this client *is*
> being phased out.

You mean this table?

http://fah-web.stanford.edu/cgi-bin/main.py?qtype=osstats

It is interesting indeed. What it tells me, however, is that only 10% 
Windows machines use GPU for computing. Together (ATI+nVidia) those GPUs 
deliver four times total PS3 power in x86-TFlops, while the number of 
units in both groups roughly matches.

I don't contribute to any of those efforts at the moment. There were times 
when I thought about it, but misc "events" in life prevented me from 
making a step. Besides, usually I need cpu too. Maybe in the future, but 
it is interesting that they have no GPU client for Linux, it seems? Too 
bad, even thou my security-based objections would make it problematic to 
run, basically, a remote execution on my own hardware. I guess it's 
perhaps good I never joined, because it would take me ages to make secure 
setup and in effect I wouldn't have contributed much anyway :-).

Anyway, I have only rough and simplistic understanding of how this all 
works.

Perhaps older clients will be able to perform some job related to 
old-style projects, while newer-faster ones will get more advanced 
abilities and will work on new projects?

So it's not all that bad. Unless you'd like to upgrade your PS3's 
software.

> Even worse, the number of processors involved seems steadily
> declining, to a point where in terms of aggregated flops, which at a
> time were bordering on the 10 Petaflops, defections are not even
> compensated by the increasing processor performance.

Generation change? I guess there was never any kind of research about 
people involved (age, education, income, likes/dislikes)?

Maybe if Falsebook told the crowd to go and install software, this would 
improve things a little.

Maybe in the future there will be Lottery at Home, when users contribute 
and the more they do, the bigger chance of winning holidays on Alaska or 
something.

> - failure to counter the general paranoia about energy (see the
> ridiculous stories about electronic devices standby...), where
> possible contributors feel more righteous in, and expect to save some
> money by, switching off their processors rather than paying the
> equivalent of a coffee a week to contribute their processor cycles for
> a better future, or at least for a better understanding of fundamental
> biochemistry.

Yup. The paranoia is very strong and multifaceted. It goes around cutting 
as much power usage as possible, while not stating the reasons behind the 
whole drive. Or giving some fairy tales for explanation.

Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature.      **
** As the answer, master did "rm -rif" on the programmer's home    **
** directory. And then the C programmer became enlightened...      **
**                                                                 **
** Tomasz Rola          mailto:tomasz_rola at bigfoot.com             **