Distributed Playstation

withinavoid writes "News.com has a story up about the next generation Playstation 3. Apparently the game developers are asking for a 1000 times performance increase and that's just not possible, so they are looking at distributed computing as a possibility. "

Comments

Playstation 3? It's already out!!

[Indras]

See here, if you don't believe me: http://www.misinformer.com/archive/01-01/15.html

Atari

[Decimal]

Distributed console computing... interesting! Does anybody have a link on how I can hook up my Atari 2600 consoles for something like this?

Just think how much faster PONG will run!

Fantastic!

[BoBaBrain]

Does this mean we can look forward to playing "CORBA Command"?

How the dialogue really went...

[gowen]

Apparently the game developers are asking for a 1000 times performance increase...

Developers: We want a 1000 times speed increase

Sony: Would you settle for a press release containing a bunch of buzzwords

Developers: Which ones?

Sony: Let me think: "distributed computing", "biotechnology", "linux", "grid computing" and "Moore's Law"

Developers: OK, if you throw in some hookers at the next Comdex in Vegas

Sony: Deal

Re:How the dialogue really went...

[poemofatic]

I think in this case, the phrase is "Moore Hookers".

Sony Hype Machine

[Spankophile]

Hmm, the last time Sony announced their "Next Generation Console" (aka PS2), it was still over a year away, while the PS1 was still selling well. The real kicker (aka purpose) though was that it was enough to keep a large number of gamers away from the DreamCast, which was a great system.

It would appear to me that hyping a PS3 while the PS2 is still selling strong would be an attempt at keeping people from getting an Xbox or GameCube.

Editorial math?

[skippy5066]

Forgive me if my math is off, but if Moore's law states that processing power roughly doubles every 18 months, wouldn't a 1000-fold increase occur in about 15 years?

Maybe they're designing the next generation Playstation on a Pentium machine. Did they ever fix that bug?


-Jeff

Re:Yes it is possible...

[Dead Penis Bird]

If you want modularity, just buy a PC.

Why does everyone want to turn game consoles into PC's? I enjoy the simplicity of the modern console game; just pop in a cartridge or CD, and play. That's it. No sysfiles to configure, no add-ons to buy (at least necessary to play most games, the N64 had a memory upgrade to play certain games, most notably the latest Zelda release).

I just wanna play dammit!

How would this work?

[martinmcc]

Its always good to see technology being pushed, but I really can't see the need for '1000' times the power for games. There is so much untapped power in the current generation consoles at the minute - compare early playstation games with the most recent, tekken 1 comparred to tekken 3 for example. In a year or so time when developers have much more experience with the hardware, I expect to see the same sort of leap. developers wanted the hardware sped up so much just sounds to me like laziness.

My other concern is how would they achieve the distribute network. The thing I like about my consoles is that I stick in the disk/catridge and play, no pissing around. I hack on my PC, I play on my console, and thats the way I like it. If I have to start into configuraing and debugging (which as it gets more complex is bound to happen) then the whole reasn for the console goes out the window.

it's not the net, dude.

[s4m7]

I've heard way too much talk about consumer applications of distributed computing lately. The trick is they are not really consumer apps at all, but merely a front. If you have a distributed network with an installed base of 10 Million machines, that's a lot of idle time you can sell off to other companies. And you can bet that that's exactly what Sony has in mind.

Think about it: the memory bandwidth on the PS2 is at least twice as fast as the FASTEST network connections available. That's nothing to say for the pitiful 2-3Mb cable pipe that is available to most people.

The graphics are what the developers want to see the improvement in. I don't think you are going to see any improvement in performance in this area brought about by distributed computing. If it's possible, I'm really curious as to how.

Re:it's not the net, dude.

[Alsee]

I think the intent of the distributed network is to handle background events in a large shared world. I describe it in this post.

On the other hand when someone is playing a multiplayer card game for example, 99.9% of the cycles and network connection are going to complete waste. It would make sense to use the excess to sell distributed processing. It could eaily make the game network a free service rather than a pay-for-play service. Heck, they could potentially let you earn credit of some sort. Leave your PS3 hooked up to the network during your vacation and come home to a pre-release coupon. 10% off on a hot new game - and get to start playing it 2 weeks before it is even available in stores.

how are they going to sell these cycles to other companies? especially since these cycles would not be sony's to sell - but mine.

Whenever you play a game they have full control of your cycles. Whenever you connect to their netork they have full control over all the data you upload and download. They could do anything they like without telling you, but it would probably be safer for them to include some wording in the licence about it - "by connecting to our game network you agree to receive, process, and transmit distributed network data".

if the memory bandwidth is so great, yet the network bandwidth is so pitiful - how does that make it a suitable topology for distributed sellable bandwidth/cycles?

For graphics you are moving huge amounts of data, but you have to update the screen several times a second. A distibuted network is useless for this kind of data.

For some projects you only need to send/recieve a few hundred or a few thousand bytes, but they can take an hour to process. Seti signal analysis or molecular protine folding problems for example. Distributed networks are great for these kinds of problems.

-

Doesn't make sense...

[Ami Ganguli]

Unless I'm misunderstanding something about the article, this makes no sense at all. Rendering a video game isn't nearly the same kind of workload as rendering a movie. The former requires low-latency, whereas the latter can be farmed out and done in batches.

There's no way you're going to get a 1000x performance boost by distributing a video game over the Internet.

I would bet that the real idea is to build in support for distributed multi-player games, and somewhere between the engineers and the marketroids things got horribly twisted.

Dumb

[rho]

I got as far as "maybe the Playstation 6 or 7 will be based on biotechnology", or some such garbage.

Please. This story is nothing more than a trumped up press release targetted towards the Xbox and GameCube in an attempt to either 1) slow their sales or 2) engender positive mindshare for the Playstation.

Distributed computing? In other words, "imagine a Beowulf cluster of these..."

I can see it now..

[dimer0]

If you wish to run Gran Turismo 4 in full resolution with highest graphics settings, 4 Playstation 3's are recommended.

Well, I guess if they're rack-mountable, I'm game. Bring it on.

Re:distributed?

[HanzoSan]

Broadband? We would need fibre optic, gig per second kinda speeds.

Also even with these kinds of speeds, how would you keep a game in sync? what about errors? how would you save? distributed too? What would keep me from cheating by worse, using the PS3 to bypass the SSSCA law and use it as a computer to share my mp3s?

Distributed computing

[jdavidb]

That sounds like a practical solution. I'll just buy a beowulf cluster of PS3's and ...

(Do these guys think I can offload the processing for my games to someone else's PS3? Won't that PS3 be busy trying to run someone else's game?)

Chip MultiProcessors?

[Erich]

Generating fast traditional processors is getting harder and harder to do. Look how fast a P4 is compared to a P3 or a P2, in terms of actual performance per transistor count. It sucks. In fact, per transistor count, smaller, simpler chips (386) do better. Since most of the performance improvement in chips comes from process migration instead of architecture (386s would run a lot faster in a .13 micron process...) one idea is to put a bunch of simple processors on a single chip.

There are several problems with this. Memory bandwith, power consumption, etc... but the main one is that most normal applications are written for a single thread.

Imagine how many MIPS 4K cores you can fit in 300mm^2 in 4-5 years. That's a lot of power. Sure, they might only run at 1-2Ghz, but there will be 64 of them on a die. If you can harness that power, it might give your game developers much of that huge performance boost they want.

Think beowulf-cluster-on-a-chip. As with multiple-workstation distributed computing clusters, the trick is not in setting the thing up, but in figuring out how to distribute your work.

Re:Chip MultiProcessors?

[chrysrobyn]

Imagine how many MIPS 4K cores you can fit in 300mm^2 in 4-5 years.

AAAAAAAH! This is my livlihood. 300mm^2 makes me scream. If you think the average consumer will be able to afford a game console that has a CPU that's 17mm on a side, I want your credit rating. Are you related to the guy whose name appears to be "Object of Envy"?

Half that size isn't bad. If you had mentioned about cramming CPUs into 150mm^2 or even 100mm^2 (I think the Game Cube processor is below 50mm^2), that would have been more realistic.

PS9

[llamalicious]

Come on... PS9 was already released, with the telepathic interface. Surely it's 1000x faster than PS2

Don't any of you watch T.V. ????

Comes from current PS2 architecture

[iamr00t]

Basicaly, we all know that it was hard for PS2 developers to make games for it initialy.
The reason was the Emotion engine in PS2, that it explicitly multithreaded, i.e. you have to make your program use all threads (unlike PPRO for example, where CPU does it for you).

It's really a whole new way to program.

Now it seems that Sony convinced some developers to lean it there's nothing stopping them from making more threads (there are 16 in Emotion if I am not mistaken).

Oh, and it has nothing to do with distiributed computing over the Internet. The application architecture is similar, but that's it. And yeah, no batches here :)

As for IBM involvement, here is the article in Wired Magasine about their cell computer

Oh, and ahother one about PS2 and PS3, that one is quite old, but explains where Sony is going.

Re:Comes from current PS2 architecture

[voronoi]

Err, the reasons the PS2 is hard to work with is:

1) Poor tools
2) Arcane DMA alighnent issues
3) Misguided selection of VU integer instructions
(no imul, you have to jump through hoops to do xor, only 16bit, yet flags are in the upper 16bits)
4) Hard to singlestep the VUs
5) Very limited blending modes in the GS
6) Very limited iterator precision in the GS
7) No hardware clipping
8) Wierd GS rules where rendering horizontal triangles is much slower than large vertical ones
9) Non perspective correct iterated vertex colour
10) Limited vram
11) 1.0 * 1.0 == 0.999999999 in the VUs

I.e. it's cheap and flawed (but hey that's a challenge and some people seem to like it)

Then you go

"It's really a whole new way to program."

You are not a programmer are you? Didn't think so.

"Now it seems that Sony convinced some developers to lean it there's nothing stopping them from making more threads (there are 16 in Emotion if I am not mistaken)."

What are you talking about? I can't possibly imagine why you would want 16 threads in PS2 game.
Genrally CPU multi threadding is quite costly. Most games are written to run at a solid 60fps*, so you can often get away with out multithreadding
stuff like the AI, the Renderer or some trickle loader.

BTW The only console that really had seriously multi-thredded games was the N64.

* Due to field rendering you have 2x the fillrate (and more vram) at 60fps than at 30fps. Dropping to 30fps is bad!

Sony has some problems.

[Iron Chef Japan]

Well I was excited about all the Cell development and this PlayStation 2 stuff, but Ken Kutargai (the guy behind playstation and SCEI president) recently made some very grim statements at the South Korean PlayStation 2 launch. On the topic of PS3 Kutaragi-san said "Nothing has been started yet." He made some very grim statements about online gaming too saying; "If broadband connections capable of delivering 10Mb/s are affixed to game consoles, the industry as we know it will be over. By that time, perhaps 2005 or later, games would be available for download rather than sold in stores." This news came right after many analyst's came out saying how skeptical they were about Sony's online plans. This comes right after the Nintendo-Square and Nintendo-Capcom deals, which by the way Kutaragi mad, summoning top Square officials to the SCE headquarters to explain the deal, as he was out of town when the deal was made (the Square one) and had no prior knowledge about it. The memory card shortage doesnt help much either.

Re:How the dialogue SHOULD HAVE gone...

[RFC959]

Developers: We want a 1000 times speed increase

Marketing guys: Yeah, and we want a 1000 times raise and an office 1000 times as big. Get bent.

Hardware guys: *kick the developers in the nerts* Give us games 1000 times better and we'll think about it, you pathetic freaks.

Very interesting Microsoft quote

[Jeppe Salvesen]

The lead developer for "Microsoft's Xbox Advanced Technology Group", Pete Isensee, said something interesting:

"Microsoft has this stigma about not getting it right until version three. We didn't have a choice with Xbox. If we didn't get it right with version one, Sony and Nintendo would eat us alive."

What is the implicit message? I would say : "As long as we have direct, real competition, we will produce quality products on time"