Slashdot Mirror
More PlayStation 3 Grid Computing Details
gwernol writes: "Redherring has a good article on Sony's forthcoming PlayStation 3: not too many technical details but good background to the Xbox/PlayStation wars. Sony are touting the use of massively parallel 'cell computing' to get a 1,000 times performance increase over the PS2. This plan, also known as grid computing is also discussed here."
Putting 1000 computational "cells" onto a chip is not difficult considering as we should be into the hundreds of millions of transistors a chip by then.
The difficulty is how to most effectively organize and utilize these units.. just imagine a 1000 machine clustor but on a chip, you've still got the same problems to sort out.
Seriously, I'm not kidding.
Though graphics in general have improved a whole hell of a lot more, they still aren't perfect. One interesting area of research is that of how the reflections and light absorption of one object affects the lighting of another/all others in a certain area. Processing power at the moment hasn't yet been sufficient to simulate the global effects caused by local objects. After all, if I have a shiny red ball next to a shiny green ball, then depending on where I put the light source, the red might partly affect the shade of the green, or vice versa. Multiply this by a factor of 10, and you can see why you might need a lot of processing power to simulate this.
Of course, this is assuming that the article really isn't just tossing buzzwords around.
Note to M1-ers: a curt but otherwise insightful message is not "Flamebait" or "Troll".
There's no point in pretending to argue against any of this because it doesn't say anything.
This is like being presented with the statement 'in ten years, men will fly like birds!'
You just can't say it's not true because it doesn't particularly carry any meaning in the first place.
When will you learn that the inherent assumption when describing a console is non-upgradeable/fixed platform/known hardware?
Making it significantly upgradeable makes it not a console.
As my father lik@(munch munch)...
Its not quite as dumb as the Redherring article makes it seem. Whoever the idiot journalist is who wrote that (not even worth my time to look it up), he obviously got confused between different forms of distributed processing. The PS3 isn't going to use SETI-like distributed processing, its going to use a large collection of local processors to do its work. Somewhat like the old Thinking Machines cubes.
Actually, the Playstation 2 is capable of 75 million poly's a second. They were all drawn at 0,0,0 with a size of 0, and the buffer they were drawn into wasn't drawn on the screen.
I shit you not, that's how they came to that #.
Hmm. That's very odd.
The PS2's geometry engine is a 4x4 arrangement (16 pixel pipelines in total), so the fastest possible render is irrespective of 4x4 or 0x0. Given that a 4x4 triangle at least possesses triangular nature, I'm surprised that they would go for 0x0.
The PS2 also doesn't use co-ordinate space of 0,0 to be anything special - the hardware has automatic scaling from an abstract 4096x4096 space into whatever resolution you happen to be working in. Typically (at 640x480), the co-ordinate (0,0) is at (1728,1808). Why then does it matter where you render the triangle ?
Whether you draw the buffer to the screen or not is also not relevant - you can't "draw" to local memory - it's drawn to RAM with a 2048-bit bus on the same chip as the video processor. There's no reason why displaying the screen would slow that down, so why open yourself to criticism if that were the case. Odd behaviour to say the least...
In short, I think you're wrong.
There are detailed figures for different types of draw operation in the GS users manual. The 75 million/sec is for no-texture, flat-shade, no-anti-alias. It drops down to 16 million/sec for textured, gourard shaded, fogged, anti-aliased triangles.
Simon
Physicists get Hadrons!