PS3's performance and Sony's long-term goal = Uber-PS3?

[Shifty Geezer]

I was thinking yesterday of the cost/performance ratio and long term goal relationship. We've all talked about PS3's specs and the ridiculous costs in the dream machine = 2 Cell, 512+ mb RAM etc. That's prohibitively expensive and would lose a LOT of money.

But, as PS2 and PS1 has shown, once you scale down the tech you can save a bundle, and then turn a profit. So, the longer PS3 is around, the more chance of it making Sony more money, right?

This then gives a possible incentive for a long-term system. Rather than release a machine at loss, after a few years it becomes profitable, and then 5 years from lauch replace it with a new loss-making system, release an uber-console that'll last 10 years, providing many more years of profitability? Instead of lots of small steps, take a giant leap.

I imagine a dream PS3 would be able to hold it's own for a LONG time. Graphically it could manage offline quality which, with more polygoins and shaders from later tech, will still be visually comparative. ie. If you produce FF's offline renders in realtime, you can't get much better until a generational shift in graphics. Such a dream PS3 would handle as much physics and AI as you could cope with, that a later system with more power wouldn't have any advantage. When you can handle 20,000 boxes colliding, a system capable of 40,000 boxes won't be able to add to gameplay. Just as PS2 sells now, even on outdated tech, because it offers a quality people are happy with it's a great earner for Sony. I imagine a dream PS3 would be even better. It won't date anywhere near as much and can sustain long term takings.

Yes? No? Discuss...

 

[aaaaa00]

Yes? No? Discuss...

No.

As long as tech continues to provide ways to increase performance at an exponential rate, nothing you can build today will stand up to anything built 5 years from now, let alone 10 years from now, not for any reasonable price.

Even xenon/PS3/rev are still going to be so far from real offline renders that there's no way they're going to be able to hold their own for 10 years.

 

[Acert93]

No way.

Physics are just becoming part of the game; by 2011/2012 not only will have graphics inproved, but so will have AI, Physics, and surely other forms of media and interaction will be available.

Think, if by 2011/2012 they are able to put 8-16 CELLS on a single core, that us 8-16x as much power. You are looking at 2-4TFLOPS from the CPU alone--that will wipe the floor with the PS3. Furthermore, the same can be applied to the graphics chip. Not only will it be more powerful, it will use better/more effecient techniques.

And memory... 256MB of memory is not gonna last long. Just based on the fact the last 2 gens increased 8 fold, by 2011 we should expect systems with 2GB of memory. And by 2001 we should be getting flash like memory devices in the 10s of GBs for very cheap, so storage will be a consideration.

And finally Sony's competition wont be sitting around doing nothing. If Sony had a 2 CELL core PS3 with 512MB of memory and hoped for a 10year life cycle--what happens when MS releases their system that has a CPU 4-8x more powerful, graphics that are 20x better, and 10x the memory--with backwards compatibility?

When you put together all the facets of home entertainment--software, new media formats, new directions in gaming, new input/output requirements (VR, camera based VR, holograms, etc...), and the expanded role of these systems I doubt we will see a company make a "10 year" product any time soon.

Another thing you need to remember is that unlike most people on this board, the 80M PS2 owners are NOT going to buy PS3s right when they come out. PS1 software sold well for a couple years after the PS2 came out. The same will happen with the PS3. So while the boards here may say otherwise, on a successful platform the release of a new system does not harken the death blow--especially when you consider poorer countries that are getting current gen stuff now. The PS3 will be around for a long time, but there is nothing that can slow the rate of progress.

If it makes you feel any better, CELL and the nVidia GPU give Sony a very clear path of a "Platform" where PS4 will not only be backwards compatible, but my guess is that developers will have a smooth transition. They will just have MORE CELLS to play with--which should make them very happy.

 

[leechan25]

Yes.

Tech life cycle will become longer the closer displaying content copy reality.
all parts of the gaming industry from content development to hardware usage are becoming longer. It fact, I won't be shock if PS3 and other consoles have greater forms of scalability like memory upgrade or chips swaps. Cell processors really lend to this idea. Maybe when PS3 is nearing the end of it's life cycle Sony will come out with some 32x like expansion increase it life.

You know you'll buy it

 

[darkblu]

i believe sony are in for pulling a wintel here. they don't have to produce anything uber - in 5 years from now they can come up with a 4-8 cell configuration running at 8-10GHz, with 1GB of ram. and if the ps3 titles are actually all opengl/es and Cg on the gpu side, ps3 emulation on the ps4 will be a grad-student-course-project-complexity task.

 

[leechan25]

Cell ability for grid processing and scalabile architecture give me the hope of a longer life cycle than past generations. No other consoles were built with these options in mind.

 

[jvd]

There is no reason to put that much power into the machine. Sony just has to put more power into it than ms puts into the xenon . That is it .

Thier big push is bluray and bluray is going to cost the ps3 alot in terms of power

 

[GwymWeepa]

I thought the inclusion of a single Cell was essentially Sony trying to win the hype war by such a large margin, that they could silence their competition once and for all. Who knows, perhaps if Sony suspects MS could win a lot of market share, they'll slap another Cell or GPU to achieve the same goals and be alone in gamer's living rooms.

But there's no way that'll happen.

 

[Phil]

I think the life cycles of the consoles are already expanding. The PlayStation launched in december of 1994 and was replaced by its successor in march of 2000 (rougly ~5 years). We already are in 2005 and have already passed the 5 year mark in which the PS2 is on the market. With the launch of the PS3 at the earliest by fall this year would make it nearly 6 years already. If Sony does things right and has such a high market share next generation, I could see it increasing to even 7 years or a bit more.

 

[aaaaa00]

I wouldn't be suprised if Sony still sells the PS3 in ten years -- the argument was never about that. I'm sure the PS3 will enjoy a long lifespan, at the end of which it will be a budget chipset integrated into toys or what not.

But I do not believe it is at all possible that the PS3 will stand up to anything built in 10 years performance-wise or in graphics.

Adding CELLs doesn't count, since a different configuration of CELLs is by definition, not a PS3 -- the PS3 is what Sony ships in 2006, not in 2016. PS5 might be backwards compatible with PS3, but a PS3 is not a PS5, just like a 386 is not an Athlon64.

 

[Legend]

Sony is planning to make PS3 last long. They said PS2’s life cycle will be ten years. So even as the nex-gen system will comeout, Sony and the others will still support the previous systems. Sony probably more than the rest- you don’t just throw a 85M user base out of the window. Heck, PSOne is still manufactured.

And about better quality; I really rather wait till I see “realâ€￾ difference as apposed to seing a slightly better system. I think 5 years will be too short for selling toasters, let alone a consumer electronics product.
I believe when PS4 is released, PS2 is still being produced
….I wonder how much will it be then. Hmm…

 

[talyn99]

I say have a hideen drive bay in PS3. WHen they shrink the chip making process, let everyone buy an add-on without having to spend a crapload of money. And an eye-toy utilizing holgraphic interfacing for games.

Not far-fetched, N64, had an expansion RAM bay. This woudl be way beyond, your NEXT console with your existing one.

 

[Teasy]

Heck, PSOne is still manufactured.

I'm pretty sure its not still being manufactured. I remember an announcement sometime in 2004 about Sony stopping manufacturing of PS1 for good.

 

[jvd]

I say have a hideen drive bay in PS3. WHen they shrink the chip making process, let everyone buy an add-on without having to spend a crapload of money. And an eye-toy utilizing holgraphic interfacing for games.

Not far-fetched, N64, had an expansion RAM bay. This woudl be way beyond, your NEXT console with your existing one.

haven't we all learned from sega that system upgrades fail .

Sega cd , 32x . On the saturn the modem ,

Nintendo

4 meg upgrade , canned cd drive

Turbo graphics

Cd drive

 

[one]

Heck, PSOne is still manufactured.

I'm pretty sure its not still being manufactured. I remember an announcement sometime in 2004 about Sony stopping manufacturing of PS1 for good.

Really?
http://www.scei.co.jp/corporate/data/bizdataps_e.html
Cumulative Production Shipments of Hardware / PlayStation®
2004/03/31 99.72 million units (Japan: 20.62 million/ USA: 39.49 million/ Europe: 39.61 million)
2004/05/18 100 million units (Japan: 20.72 million/ USA: 39.67 million/ Europe: 39.61 million)
2004/06/30 100.29 million units (Japan: 20.77 million/ USA: 39.91 million/ Europe: 39.61 million)
2004/09/30 100.89 million units (Japan: 20.95 million/ USA: 40.19 million/ Europe: 39.75 million)
2004/12/31 101.73 million units (Japan: 21.13 million/ USA: 40.78 million/ Europe: 39.82 million)

 

[randycat99]

Yes, we've seen upgrades fail as a matter of precedence. However, we've also seen Sony break a lot of precedences.

Personally, I agree there certainly is a lot of reasons against such an outcome, but I wouldn't rule it out completely. If the time has come for something like that to happen, more power to it...(no pun intended) I'm more than curious to see how it turns out. I don't think it is a cosmic rule that we need new generations of equipment on a steady, regular basis. When the game designs are really calling for more power that cannot be done any other way, that should be the impetus. Until then there is always plenty of room for developers to tweak and optimize to accomplish what they want.

 

[Megadrive1988]

I was thinking yesterday of the cost/performance ratio and long term goal relationship. We've all talked about PS3's specs and the ridiculous costs in the dream machine = 2 Cell, 512+ mb RAM etc. That's prohibitively expensive and would lose a LOT of money.

But, as PS2 and PS1 has shown, once you scale down the tech you can save a bundle, and then turn a profit. So, the longer PS3 is around, the more chance of it making Sony more money, right?

This then gives a possible incentive for a long-term system. Rather than release a machine at loss, after a few years it becomes profitable, and then 5 years from lauch replace it with a new loss-making system, release an uber-console that'll last 10 years, providing many more years of profitability? Instead of lots of small steps, take a giant leap.

I imagine a dream PS3 would be able to hold it's own for a LONG time. Graphically it could manage offline quality which, with more polygoins and shaders from later tech, will still be visually comparative. ie. If you produce FF's offline renders in realtime, you can't get much better until a generational shift in graphics. Such a dream PS3 would handle as much physics and AI as you could cope with, that a later system with more power wouldn't have any advantage. When you can handle 20,000 boxes colliding, a system capable of 40,000 boxes won't be able to add to gameplay. Just as PS2 sells now, even on outdated tech, because it offers a quality people are happy with it's a great earner for Sony. I imagine a dream PS3 would be even better. It won't date anywhere near as much and can sustain long term takings.

Yes? No? Discuss...

I disagree for the most part. except for that, I would like a 'dream-PS3' myself, but my dream-PS3 probably far outclasses what you are thinking of.

anyway, even the most powerful PS3 that is within the realm of possibility from a technological standpoint (not price standpoint) would still NOT be able to do highend prerendered (offline) CGI graphics. so, that means that your 'dream-PS3' would be able to do less, and, the real actual PS3 that is coming out will be doing even less still. not that the real PS3 isnt going to be powerful and impressive, I think it will be.

now, if you are talking about older CGI from Playstation1, games, that is a possibility. but not current videogame CGI much less feature film calibur CGI.

the best we could hope for is probably Final Fantasy 8, Tekken 3, Final Fantasy 9 CGI. and even that might be out of reach. Im really not sure.

and there will most likely not be any upgrades to PS3. the only thing that might be remotely possible, is linking PS3s together, physically (not over the internet) like sort of an SLI. but even that is doubtful.

by 2012 or so, we'll have PS4, and it should do everything that we had hoped PS3 would do in our wildest dreams. but still not everything we'd want in the long run, thats why PS5 will be made sometime in the later part of the next decade, 2018-2019 with maybe optical processors and real raytracing. PS4 in 2011-2012 should be another Cell-based system with 100 or so cores.

anyway, I believe that the real PS3 will be able to do the things that were *reasonably* expected of the PS2 (not the toy story or film quality graphics but stuff like Final Fantasy 8 CG) ) but could not do.

 

[talyn99]

the add-on idea is Just right for this generation, and given Cell's scaleablility. When they shrink the chip making progress, stuff a more shrunk down expension board. Increase performance dramatically.

Yeah, I would say the Sega CD, is not comparable to this generation;s flexibility. Come on, how far you think graphics will jump ahead, processors this time eclipsing GPU's marginally i feel. Unless the PPU's makes a real difference. Well, DNA computing, I have to read up to see hwo far along they are with that. But how many of you think the graphic cards by end of five years will only be 3-4 times stronger thanthe upcoming crop of consoles.

peace

 

[Megadrive1988]

unless we see a shift in policy by Sony, Nintendo and Microsoft, we will not be seeing upgradable consoles next generation. even typically upgrade-happy Sega did not make the Dreamcast upgradable. not talking about *expansion*, I am talking about *upgrades* as far as CPU, GPU and RAM.

 

[Shifty Geezer]

I forgot to mention the other part of the argument, which is ending the competition once and for all.

Firstly an uber-PS3 will have a numbers-edge over Xenon for marketting, even if first gen titles don't look much difference. Secondly, 3 years down the line when the PS3 hardware is maybe nearing breakeven cost point, it'll be showing it's strength. By adding uberpower, MS looks like smallfry, competing only on price. Within four years, XB2 could be dead in the water, shifting no more units, giving PS3 clear cruising for another dew years while MS, if they haven't retreated by then, to research a new system. It's not just about being in the number one slot, but forcing MS out of the market.

Though undeniably tech will be leaps and bounds ahead, performance of a system isn't just limited by release time but also costs. At a huge loss, Sony could shovel twice as much performance into PS3 as any rational company would dare, giving it a 2 year head start, adding 2 years shelf life. Already the development of Cell is giving a larger increase to brute-force computation than other techs in the same timeframe. That is, if you extrapolate MS's approach forwards 5 years, what power difference will they have over a 2 Cell PS3?

Regards nextgen visuals, though cinema quality is out, I expect game cutscene CGs like FFX. I imagine a lot of that will be possible in realtime. Maybe not the huge cities or complex water effects, but definitely the quality of character models.

As another point, if the tech advantage that comes with a few years wait is that great, why doesn't one of the companies miss this launch window and instead launch a new platform in 2008, where, in theory, they'd be a lot more powerful (4x)?

Of course it's unlikely we'll see UberPS3, but looking at the options, long-term savings and dominance over a defunct MS, I'm suggesting there might perhaps be reason that (hopeful) rumours of 512 GFlop Cell processing, 512+ mb RAM and so forth *could* be a possible consideration, and not prohibitively expensive if you factor in the long term benefits. Bear in mind there's still that home media stuff that Sony and MS are gunning for, and with MS's strength in the PC market Sony could do with knocking out the XB franchise. UberPS3 is the most likely way to deal such a blow and gain that monopolistic percentage cut of all media that these giants seem to be after.