Another awesome statement about the SPEs of the PS3.

[makeitlookreal]

http://www.nzherald.co.nz/section/story.cfm?c_id=5&objectid=10462728


Tell us about the overhaul of the engine technology to take advantage of the new console?

We are continuing to build our Insomniac Engine and have made many improvements to it since Resistance: Fall of Man. The one huge focus for us has been moving more of our processes over to the SPUs on the CELL processor. This has allowed us to get our physics and effects systems running roughly four times faster than it did in Resistance at nearly double the framerate, which is something you can see in weapons like the Tornado Launcher.

[Nameless]

I hope true third party developers jump on board...

[Viper]

I hope true third party developers jump on board...

I assume you mean you want more exclusive support? If not, what do you mean by true 3rd party developers?

[masteratt]

Ones that instead of whining about the PS3, who work properly on it and deliver a solid running game?

I'm glad to hear this.
120fps here we come!! (kidding, don't you dare take this seriously).

[makeitlookreal]

That is the big issue this generation. Those developers that are willing to spend the energy, time, and money on the PS3 can get amazing results. Those that are not willing to do those things will get mediocre results.

The CELL processor is the key to the PS3. It can do so much awesome stuff! However, it's more complicated than what many developers are used to. But if they are willing to do the work they can do amazing things.

I think the problem is that many developers just are not willing to spend the time, energy, and money to get the most out of the CELL.

[Viper]

I think the problem is that many developers just are not willing to spend the time, energy, and money to get the most out of the CELL.

You nailed it here. Will they get their money back?

Think of it this way, they are spending nearly 10 times as much to make the games but the price of the games are up only $10.00 over last gen and the install simply cannot compete with the luxury the PS2 afforded.

In a crude sense, we asking them to lose a lot of money so we can have the prettiest games.

Publishers are being forced to find a happy medium between what they can afford to spend and what they think they'll make back in return. The only way to break that cycle is to gain a much bigger install base but how do you get that? Lots of great software. The cycle comes right back.

It's very precarious right now but so long as publishers balance things enough and try to recover losses with downloadable content, in game ads and quick buck cash ins, they should be ok.

[Segitz]

Well Viper, keep in mind that, in effect, the only major difference in making a game for the PS3 compared to the 360, is the CELL. So the rest, assets, sound etc. etc. etc... is definately the same.

So, are engine programmer so goddamn timeconsuming/expensive or don't the publishers give them enough time to complete their work? I think it is the latter, especially with EA and its yearly crap licences.

I am not asking for every game to look like or better than MGS4, but it at least should ran the same as on the 360, since we KNOW PS3 is in every way as potent as the 360 (maybe not in AA, but this does not result in 30FPS to 60FPS in a freakin madden game).


This stuff is easily remedied next year, when the PS3 had enough airtime with the CELL so the devs should have learned. Since I never buy sports games (my brother usually does once per generation, I still have FIFA2001, man that graphics suck big time, essentially PS1 graphics, no kidding, even the menus look horribly low res!), I do not care much, but the bigger mainstream crowd does (but since the usual casual buyer is as dumb as shit, they won't even know I think).

[cpiasminc]

So, are engine programmer so goddamn timeconsuming/expensive or don't the publishers give them enough time to complete their work?
Yes, and yes. The big time and expense of engine development isn't the engine itself, but making it useful. Moving physics to the SPUs isn't a time-consuming job, but making a pipeline that can allow artists to bring stuff into the engine and making things repeatable, testable, trackable, and verifiable, while getting through all the game-specific bugs and locally BS-ing things... all that is where the difficulty lies, and when you're working on an actual game in production, it's not the SPUs that will get priority, it's supporting the game team (and that's how it will always be).

Teams like Insomniac are able to make it happen because they're only on one game at a time. Their current project was likely in the midst of production transitions at the time they were going through all that. Nobody is blocked from trying to push their engine up.

Also, some of those numbers sound like per-SPU-job results, which always sound much better than the overall. Getting an 8:1 speedup per job doesn't mean the effects system runs 8x faster or can move 8x the effects.

[makeitlookreal]

You nailed it here. Will they get their money back?

Think of it this way, they are spending nearly 10 times as much to make the games but the price of the games are up only $10.00 over last gen and the install simply cannot compete with the luxury the PS2 afforded.

In a crude sense, we asking them to lose a lot of money so we can have the prettiest games.

Publishers are being forced to find a happy medium between what they can afford to spend and what they think they'll make back in return. The only way to break that cycle is to gain a much bigger install base but how do you get that? Lots of great software. The cycle comes right back.

It's very precarious right now but so long as publishers balance things enough and try to recover losses with downloadable content, in game ads and quick buck cash ins, they should be ok.

Viper,

That is why games should not all cost the same price. For example, I think a game like Killzone 2 that's going crazy with SPE usage (and cost a lot of money to produce) should sell for a higher price than games that took much less effort or cost to produce.

[LiquidEagle]

That's already the case sometimes -- Katamari Damacy was a cheap game to make and sold for $20 new when it first came out... It's not all gaged on SPU usage though...

[Jay Gee]

Viper,

That is why games should not all cost the same price. For example, I think a game like Killzone 2 that's going crazy with SPE usage (and cost a lot of money to produce) should sell for a higher price than games that took much less effort or cost to produce.
You've already forgotten the fact they plan on releasing DLC for the game to offset the development costs. Upping the price on the games for a console with an already high price tag will only further alienate your fanbase.

[cpiasminc]

That is why games should not all cost the same price. For example, I think a game like Killzone 2 that's going crazy with SPE usage (and cost a lot of money to produce) should sell for a higher price than games that took much less effort or cost to produce.
Cost and effort or SPE usage? Coming from you, I'd expect you to consider those to be an absolute equivalent... It really wouldn't take much to have a game of far higher cost, effort and production value than Killzone that doesn't use a single SPE.

The real question is what kind of price range variance the market will bear, and it's typically really small. Budget titles sold at budget prices is not a concept I'd say is a good idea in large numbers. As an occasional thing here and there, it's harmless, and has little impact on the market overall. Often times, it works out, but this happens at the hands of larger publishers who can afford to do that sort of thing. $5 million budgets would be considered "on the cheap" for an EA, but for, say, Atari, it'd be hanging on by a thread.

Fairness is all well and good, but you're trying to sell what is effectively a luxury item -- self-centeredness becomes the hallmark of the consumer, and fairness goes out the window.

[Viper]

Well Viper, keep in mind that, in effect, the only major difference in making a game for the PS3 compared to the 360, is the CELL. So the rest, assets, sound etc. etc. etc... is definately the same.

So, are engine programmer so goddamn timeconsuming/expensive or don't the publishers give them enough time to complete their work? I think it is the latter, especially with EA and its yearly crap licences.

I am not asking for every game to look like or better than MGS4, but it at least should ran the same as on the 360, since we KNOW PS3 is in every way as potent as the 360 (maybe not in AA, but this does not result in 30FPS to 60FPS in a freakin madden game).


This stuff is easily remedied next year, when the PS3 had enough airtime with the CELL so the devs should have learned. Since I never buy sports games (my brother usually does once per generation, I still have FIFA2001, man that graphics suck big time, essentially PS1 graphics, no kidding, even the menus look horribly low res!), I do not care much, but the bigger mainstream crowd does (but since the usual casual buyer is as dumb as shit, they won't even know I think).

You also have to understand the X360 has been out a year longer so developers have more experience with it and the X360 was designed to be developer friendly (not suggesting it's a simple to develop for either).

When you literally have to weigh your budget, system experience and install base, choosing the X360 is not a tough decision unless you want a higher sell through in Japan then you're considering a complete game overhaul and choosing between PS3, Wii, PS2 or DS.

The potential of the CELL doesn't outweigh the benefits of X360 development (right now). Of course that could all change in a year or two as the install base rises and familiarity with the system increases but given the vastness of big game development cycles it's hard to way when those X360 benefits will lapse and tapping the CELL potential will become lucrative.

[Segitz]

Yes, and yes. The big time and expense of engine development isn't the engine itself, but making it useful. Moving physics to the SPUs isn't a time-consuming job, but making a pipeline that can allow artists to bring stuff into the engine and making things repeatable, testable, trackable, and verifiable, while getting through all the game-specific bugs and locally BS-ing things... all that is where the difficulty lies, and when you're working on an actual game in production, it's not the SPUs that will get priority, it's supporting the game team (and that's how it will always be).


My post should have been a rhethoric question :D

What I meant was, is programming stuff on the SPUs so much more time and money consuming that doing stuff on the XeCPU? I guess it takes more time for trial and error if you are new to it, but as I said, next year, I doubt we'll see much difference in EA Sports games on either platform.

Seeing as such games can run at 60FPS on the 360 makes the PS3 version "broken" in a way. I mean, bad performance is also a bug, don't you agree? I know, they had to ship and they had their deadlines... In other industries (mine for instance, automotive engineering), if we ship broken hardware, we will get sued to hell (ok, our stuff can kill people :D), but here it only means worse reviews, which shed a bad light onto Sony (and less onto EA, considering the casual gamer has no idea who makes his games, really).

You also have to understand the X360 has been out a year longer so developers have more experience with it and the X360 was designed to be developer friendly (not suggesting it's a simple to develop for either).

When you literally have to weigh your budget, system experience and install base, choosing the X360 is not a tough decision unless you want a higher sell through in Japan then you're considering a complete game overhaul and choosing between PS3, Wii, PS2 or DS.

The potential of the CELL doesn't outweigh the benefits of X360 development (right now). Of course that could all change in a year or two as the install base rises and familiarity with the system increases but given the vastness of big game development cycles it's hard to way when those X360 benefits will lapse and tapping the CELL potential will become lucrative.

I said that, didn't I? I said, PS3 games will get better come they get more experienced.

I rather think, the CELL, it being much more complicated, is very much hindering PS3 development now (but later on it helps them bigtime). I couldn't care less of Madden runs with less FPS (I won't buy it) and whatnot.

[Viper]

Your broken analogy fits but only partly. Given that no one will ever get 100% performance from either system, it could then be stated every created for it is broken. Given the context of your analogy, it's more accurate to say it's more broken than the X360 version. Broken as a % of output vs 100% theoretical performance. But that's all just being anal with semantics I suppose.

The only reason I elaborately reiterated your statement regarding future CELL development is because of how future development is still relative to today. Sales trends of today and current knowledge of each system (though CELL knowledge will increase, so too with X360 knowledge) are factored into the equation I discussed in my previous post.

One other issue to consider regarding the potential is exactly that...what IS the potential vs future X360 development knowledge? Obviously developers will be able to answer that question for themselves much easier than any of us but to me it's a valid dilemma. Will CELL just give an Xbox/PS2 like difference in terms of development, something less, something greater? Depending how this question gets answered will go a long way in determining the rest of the factors value and ultimately where your dev team will be spending their next 2-3 years.

[ddaryl]

I think the problem is that many developers just are not willing to spend the time, energy, and money to get the most out of the CELL.


Of course it is. It's about profit margins.

How much $$$ do you have to put into a game to make what % of profit.

[cpiasminc]

What I meant was, is programming stuff on the SPUs so much more time and money consuming that doing stuff on the XeCPU? I guess it takes more time for trial and error if you are new to it, but as I said, next year, I doubt we'll see much difference in EA Sports games on either platform.
Well, if we're talking about eke-ing something out of the CPU and multithreading, I'd say XeCPU is the harder of the two (which is a roundabout way of saying that Win32 threads suck and SPURS jobs kick ass). The problem is that 360 doesn't demand very top-notch multithreading or scalable multithreading to get sizeable early impact (and this is partly due to the fact that it is symmetric and it is few cores). You can get pretty good results out of the 360 while not multithreading at all -- and that's the general factor over which people crow about the PS3. Not so much the SPUs themselves, but that it *forces* you to have to think about problems which are never going to be simple or easy to solve on any platform... 360 doesn't force your hand on anything, and that's why people can get their 60 fps Madden on 360 so easily.

SPU culling of polygons, for instance -- it's not difficult at all, but the fact that you basically have no choice but to do something to that effect on the PS3 (or at least be clever about how you issue geometry) is what people hate. Although, I'd add that dynamically culling every frame also means that you've destroyed the advantages of having precompiled display lists, which is pretty flat out bad if you don't get a sizeable advantage for that culling.

I'd also add that Microsoft encourages that developers do a lot of things which are just bad ideas simply because they do okay on 360 or PC, but they clearly hurt you on PS3. Moreover, when it became clear the SPURS-style job queue model works well for both CELL and XeCPU, Microsoft goes out of their way to propagandize that you should NEVER take advantage of this (and if you try to tell them otherwise, they just tell you you're doing it wrong). They try to ensure that you only do things which are bad ideas on the PS3 regardless of whether or not they're especially good ideas on Xenon or PC. Many cross-platform developers are only beginning to scoff at Microsoft's proclamations these days (hard to do sometimes since they do squeeze it in the middle of some pretty cool stuff).

Seeing as such games can run at 60FPS on the 360 makes the PS3 version "broken" in a way. I mean, bad performance is also a bug, don't you agree?
That's kind of up in the air. Inconsistency and failure to keep up performance is certainly a bug and or generic failing. But if you get consistency at 30, then that just suggests that it was designed to make 30 on that one platform... You could just as well talk about a car which has multiple trims which have different powerplants, though obviously the circumstances that create the discrepancies breaks down the analogy.

[VG Aficionado]

I'd also add that Microsoft encourages that developers do a lot of things which are just bad ideas simply because they do okay on 360 or PC, but they clearly hurt you on PS3. Moreover, when it became clear the SPURS-style job queue model works well for both CELL and XeCPU, Microsoft goes out of their way to propagandize that you should NEVER take advantage of this (and if you try to tell them otherwise, they just tell you you're doing it wrong). They try to ensure that you only do things which are bad ideas on the PS3 regardless of whether or not they're especially good ideas on Xenon or PC. Many cross-platform developers are only beginning to scoff at Microsoft's proclamations these days (hard to do sometimes since they do squeeze it in the middle of some pretty cool stuff).Wow. I think some people will find this very interesting.

[curryking1]

That's pretty crazy, not that I'm going to just start 'hating' something because of it, but that's some serious business, that's like uber tactical espionage-type sabotage.

Just goes to show how good it's is to be in the know, knowledge is definitely a strength and that's pretty clear for any side here.

[cpiasminc]

Maybe, but it's not as though I'd expect Microsoft to be unique in that respect. I can't say I've been to very many Sony developer talks recently, but I would be surprised if they didn't do similar things (though it's a little harder to find something that's a good idea on PS3 that isn't also a good idea on 360). I've heard people accuse Nintendo of worse things than that, but I would really have no clue. Nintendo is kind of in an island this generation, so whatever they say to developers is pretty separated from the rest of the universe. That said, industrial espionage is something a lot of companies are very paranoid about in this industry -- I don't know how much I buy into the idea, but there are likely some reasons for it.

It can be something of an Achilles heel that software development is something that is such a direct mapping from conceptualization to product -- anything that influences the way you think ends up being massive in the end result.

[VG Aficionado]

Maybe, but it's not as though I'd expect Microsoft to be unique in that respect. I can't say I've been to very many Sony developer talks recently, but I would be surprised if they didn't do similar things (though it's a little harder to find something that's a good idea on PS3 that isn't also a good idea on 360). I've heard people accuse Nintendo of worse things than that, but I would really have no clue. Nintendo is kind of in an island this generation, so whatever they say to developers is pretty separated from the rest of the universe. That said, industrial espionage is something a lot of companies are very paranoid about in this industry -- I don't know how much I buy into the idea, but there are likely some reasons for it.Yeah, it's fair to think that all companies would try to favor their platforms and describe not-so-optimal methods that could be used on competitor's hardware. It's not like it's a huge revelation: it's just that it's Microsoft, and that it's unsurprising.

[fastasleep]

Viper,

That is why games should not all cost the same price. For example, I think a game like Killzone 2 that's going crazy with SPE usage (and cost a lot of money to produce) should sell for a higher price than games that took much less effort or cost to produce.

Yeah, that's going to go over well. Good luck. If a movie costs 500 million to make, should we be expected to pay $50 per ticket. Guerilla Games has yet to qualify as a top tier developer even warranting a $60 price tag let alone higher.

I don't care how much they spend on Killzone 2 or how much spe utilization the game will have, if its not fun to play, they're not getting dime one from me.

[yoshaw]

I can't say I've been to very many Sony developer talks recently, but I would be surprised if they didn't do similar things (though it's a little harder to find something that's a good idea on PS3 that isn't also a good idea on 360).

CPI,

what do you perceive of the NEON engine from Codemasters? According to them, it was developed on top of SCEI's engine of some sort(I'm forgetting the name here). And there was this 2-3 page article in the digital edition of Develop Magazine in which Codemasters themselves appreciated the fact that Sony went out of their way to help build a cross-platform engine i.e NEON now.

How would you see that effort of Sony as a developer yourself? You heard anything more in this regard, please do share?

Sorry if this sounds irrelevant but I thought I might share.