On Sun, Nov 27, 2005 at 02:15:19PM -0800, [EMAIL PROTECTED] wrote: > > AI is a an interesting real-world problem. AI is massively parallel. > > Gaming is an interesting real-world problem. Gaming is massively parallel. > > So are physical simulations. > A couple of real-world game developers comment on the Cell:
I'd rather have a couple of real-world number crunchers comment
on the Cell, frankly. All game dev punks can think of is how to render
more and prettier polygons, preferrably without throwing away their
old codebase.
> Gabe Newell of Valve:
> on Sony's Cell processor ("tiny little errors can grind [a program] to a
> halt, and there is no visibility into why that's happening")
> http://www.joystiq.com/entry/1234000350057192/
"Parallel programming and debugging is hard". Well, no shit, Sherlock.
Get used to it while there's time.
> John Carmack of ID:
> So John's first impressions are, "360 great, PS3 - pain in my ass."
> http://www.eurogamer.net/article.php?article_id=60510
"Parallel programming and debugging is hard". Well, no shit, Sherlock.
Get used to it while there's time.
> http://www.eurogamer.net/article.php?article_id=60510
>
> Anandtech:
> The Cell processor doesn't get off the hook just because it only uses a
> single one of these horribly slow cores; the SPE array ends up being fairly
> useless in the majority of situations, making it little more than a waste of
> die space.
What can I say, Anandtech folks' problems (not that I'm aware them
being noted number crunching programmers) are not my problems.
Local-connectivity integer automata networks and large scale MD would
run just fine on SPEs.
> We of course asked the obvious question: would game developers rather have 3
> slow general purpose cores, or one of those cores paired with an array of
> specialized SPEs? The response was unanimous, everyone we have spoken to
> would rather take the general purpose core approach.
> http://groups.google.com/group/alt.games.video.sony-playstation2/msg/62ff83d96ea78ea9?hl=en
"Parallel programming and debugging is hard". Well, no shit, Sherlock.
Get used to it while there's time.
> I don't mean to entirely harsh on the Cell. It does some things very well,
> and the PS3 will probably do very well as a platform. But anyone who thinks
> it's a radical new improvement in computer architecture simply doesn't
> understand the problem.
There seems to be a basic misunderstanding between what one wants, and what one
gets. What I want is something far more radical than the Cell. But the Cell is
the only useful architecture I can get for a few $100/node. Current dual-core
AMD64 desktops are at about 700 EUR. If such dual-cores with DDR2 built-in will
ship for half the price by the time the PS3 Linux kit lands the PS3 will not
that interesting as a platform but for power savings (which can be considerable
over lifetime of a cluster, especially in if it's airconditioned).
I've been pretty disappointed by the PS2 in that respect. There's a fair chance
that PS3 will be no different, and that IBM's Cell blades will be targeted for
the supercomputing market and a suitable price tag.
--
Eugen* Leitl <a href="http://leitl.org";>leitl</a>
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
-------
To unsubscribe, change your address, or temporarily deactivate your
subscription,
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
signature.asc
Description: Digital signature
