Sony is officially working on PlayStation 4

[liolio]

Forget for a minute that I said launch at a lower price. Hypothetically what if Sony spent all it's RnD budget on a new GPU, a GPU that in all likelihood would have more potential than a competitors GPU who also has to spend money sourcing a new CPU (given identical budgets).
The consensus seems to be that the PS3 is unbalanced with too much emphasis being placed on CPU power. Would that CPU power (a nine core 3.2Ghz processor) still be adequate in a couple of years time?

I believe that the cell could be adequate with minor adjustments as long as Sony invests properly in the GPU. The PS3 is not really unbalanced in regard to the weight of the CPU and the GPU it's more like Ms had acces to a pretty competitive part via ATI (unified design + edram). Ms and SOny spent mostly the silicon budget for their GPU.

 

[3dilettante]

If Sony really do want to reduce their RnD spend then what are the chances that the current Cell, with perhaps the 8th SPU returning due to better yields, being used in PS4?

I suppose if they wanted to cheap out, they could.
If the overall design remains the same, this may be getting near the shrink limits of Cell. The pads for the IO and DRAM don't shrink as well per node. They may not be a limiter at 45nm, but potentially an additional shrink done in the PS4 generation could leave them with with area they could not use but would still be paying to fab.
I suppose they could aim for an on-die GPU like the latest 360 chips, but that could constrain improvements to the GPU.

 

[function]

There is no spin. It just is what it is. Why are you talking about disaster and I'm talking about forward thinking R&D concepts? BTW, how do you know MS will have "ease of development" next generation? That sounds a bit like wishful thinking to me. It may end up being the truth, but you have nothing to base that on, now. Multi-threading will be the order of the day, next-gen. Cell's practices completely carries over to the future.

R&D is supposed to tie in to the future success of the business.

How do I know MS will work hard on "ease of development" next generation? Nothing to base it on? Are you serious?

You read that just fine. Take a look at one of DICE's recent slides with one stat on culling, I believe. There are others. That wasn't an "out of the blue" statement. There are tasks that Cell is about twice as fast at than modern 4-core Intel offerings.

There are tasks where a modern quad core Intel processor is an order of magnitude (or more) faster. But cherry pick away!

[Edit]That's an estimate-come-guess btw, so I shouldn't have stated it like it's some kind of fact. There will be game applicable stuff like AI where a modern Intel Quad core could be much, much faster though.[/Edit]

Now, THAT'S "spin".

Which bit?

New first party games (big titles) don't, usually, have idle CPU time.

God of War? Killzone 3? LBP2?

They keep finding different ways to use Cell to make the GPU more efficient. That concept works in ANY home console gaming setup. Bigger GPU's won't change the fact that some burden can be taken off of the GPU to make games look better! That within itself defeats the argument of having a better graphics chip. The fact is that, at any given time, there is only one top GPU you can have from a given manufacturer.

Let me get this straight: the fact that you can use a chunk of your CPU time to make up for deficiencies in your GPU is proof that you shouldn't have a better GPU?

In order to push graphics further than what's available at that time, you need to have a stronger CPU to ease that GPU's burden with would be idle time.

Or you could just put the silicon into a better GPU!

If you're stuck with a geriatric, EOL GPU then I guess anything extra is a bonus though.

I'll bet you that next gen Sony won't be wanting a billion dollar supercomputer-on-a-chip sat inside every console, idling, though.

 

[Shifty Geezer]

Can I just remind people there's been a thread discussing all these points for the past few years. There's little point restarting the same discussions in another thread. This one provides news of Sony's PS4, which, let's be honest, isn't much by way of news. Anything directly related can be mentioned here, but speculation about what PS4 could be really belongs here with the rest.

 

[i_am_interested]

That's an incredible amount of effort you're putting in to spin this.

Sony never intended for this generation to be something of a disaster, and I doubt that RSX was actually an expensive, ahead-of-its-time investment in PS4.

MS will have "ease of development" next generation without without sacrificing this generation, and I expect the same thing from Nintendo.



What?

is this a serious post? i cant tell if it is or if it is just sony hate?

disaster? whats disastrous about a console that could likely sell 100 million? or an architecture that could find its way into possibly 200 million if spread over 2 generations?

sacrificing a generation? most studios took a game or 2 to figure everything out, and we're at the point where some studios are on their 5th or 6th game for this gen already

the only devs still "struggling" are the ones who happen to be releasing their first game for ps3 5 years into the consoles lifecycle and studios who continue to fight the ps3 by treating a console of similar install base as some type of 3rd class product (wont name any there) --- companies like id and valve solved both by tackling the ps3 head on

hopefully it doesnt blow your mind to know that some devs already find developing on ps3 easier than 360 at this point into the generation

 

[liolio]

So they won't spend as much in R&D which sounds just nice.
By looking at the PC world and the time line Sony is aiming, a pretty simple and efficient bet would be to push out a 4 cores OoO POWERPC with 256bits wide SIMD (as in PC world) and keep the 8 SPUs.
SPUs are still doing great at their job, would ensure easy BC etc.
I don't believe ARM designs will be "there" in power vs PPC and X86 by the time the system will ship.

 

[function]

is this a serious post? i cant tell if it is or if it is just sony hate?

Pack that nonsense in.

disaster? whats disastrous about a console that could likely sell 100 million? or an architecture that could find its way into possibly 200 million if spread over 2 generations?

"Something of a disaster" : losing billions that you didn't need to (money that would have put Atari, 3DO and even Sega out of the market several times over) while going from first place to last place and losing your top slot with devs, pubs and retailers.

It didn't need to have happened. It's cool that Sony have stuck with the PS3 so unflinchingly though. I have a lot of respect for that.

sacrificing a generation? most studios took a game or 2 to figure everything out, and we're at the point where some studios are on their 5th or 6th game for this gen already

Sony didn't sacrifice this generation to set Cell and RSX up for a Return of the King style super triumph next generation. That was the point in question.

the only devs still "struggling" are the ones who happen to be releasing their first game for ps3 5 years into the consoles lifecycle and studios who continue to fight the ps3 by treating a console of similar install base as some type of 3rd class product (wont name any there) --- companies like id and valve solved both by tackling the ps3 head on

Lazy devs?

hopefully it doesnt blow your mind to know that some devs already find developing on ps3 easier than 360 at this point into the generation

1) Who finds it easier, and under what circumstances?
2) "At this point in the generation" is when Sony start talking about PS4.

Being able to handle ports from PS3 to PS4 is not something Sony should focus on if they have even the remotest bit of sense. Being able to easily handle multiplatform games that also target the Xbox 3 and PC - especially early on - is what Sony should be focusing on.

 

[hoho]

I don't believe ARM designs will be "there" in power vs PPC and X86 by the time the system will ship.

I believe arm15 delivered pretty decent speeds. Not quite I7 level but definitely ahead of Atom and I wouldn't be surprised if also clock-per-clock better than PPC cores in xb360/cell. Throw a decent SIMD unit in there and they should be just fine.

 

[mrcorbo]

We don't really know what an ARM-based design that didn't have such strict power/die size restrictions would perform like or really even look like. I think some are confusing "have not" with "can not" when it comes to them. I'd certainly like to see them try their hand at a console-targeted product given what they and their partners have achieved in the mobile space.

 

[liolio]

I believe arm15 delivered pretty decent speeds. Not quite I7 level but definitely ahead of Atom and I wouldn't be surprised if also clock-per-clock better than PPC cores in xb360/cell. Throw a decent SIMD unit in there and they should be just fine.

It's possible but PPU and Px are not good examples of high performance per cycle CPU designs. I'm confident that G4 or G5 derivatives would outdo A15 in perfs per cycle (at the cost of higher power consumption that's for sure). PPU and Px have for them selves that they run at high frequency which make up for some of their lacking.
But if Sony have a lesser budget I don't how much IBM would ask to make a POWER4/5 or POWERPC 970xx derivative. IBM lost Apple in 2005 they may not have something newer than that, their other chip are pretty much server oriented. MS may face the same problem.
If neither of them want to do something that IBM may have interest in they'll have to found all the R&D. Ms got lucky last gen with Xenon as it looks like they got it for cheap.

Maybe ARM is a better option, they started to talk about the arch after A15? No I can't remember for sure. It would qualify more easily as replacement for sucky but high frequency CPU as PPu/Px

 

[patsu]

Be careful. Don't believe everything you read on the net.

Meanwhile...
http://www.smarthouse.com.au/Gaming/Industry/F5C6F8A6?page=1

What we have been told is that Sony is working on a very fast processor which is being developed with IBM, who was the original architect with Sony and Toshiba of the Cell processor, which has been used by both Sony and Microsoft to power their gaming processors for nearly 10 years.

IBM sources claim that the new multi core PowerPC processor, which Big Blue, has spent several years developing is now part of a joint development project with the Japanese company.

Sony recently moved to buy back the Toshiba Cell factory in Nagasaki for $600 million with some tipping that this will be used to manufacture the processor for both Sony Toshiba and IBM devices.

The new 32nm Cell processor is tipped to be capable of up to 16 SPEs which is twice as fast as the current Cell processor according to IBM leaks.

Japanese sources claim that Sony is gearing up to manufacture the Cell processor in bulk with some analysts tipping that the new processor will also appear in Sony notebooks and built into new Sony Bravia TVs.

Intel sources have said that they are also working on a new processor that will be ideal for future gaming consoles as well as for use in devices that are both gaming console and media hubs.

Come on...

Spoiler

interconnect !


... and improve the memory architecture.

EDIT: Dug around the net. Rumor may be false. ^_^

 

[i_am_interested]

"Something of a disaster" : losing billions that you didn't need to (money that would have put Atari, 3DO and even Sega out of the market several times over) while going from first place to last place and losing your top slot with devs, pubs and retailers.

It didn't need to have happened. It's cool that Sony have stuck with the PS3 so unflinchingly though. I have a lot of respect for that.

Sony didn't sacrifice this generation to set Cell and RSX up for a Return of the King style super triumph next generation. That was the point in question.

Lazy devs?

1) Who finds it easier, and under what circumstances?
2) "At this point in the generation" is when Sony start talking about PS4.

Being able to handle ports from PS3 to PS4 is not something Sony should focus on if they have even the remotest bit of sense. Being able to easily handle multiplatform games that also target the Xbox 3 and PC - especially early on - is what Sony should be focusing on.

they didnt plan on losing billions, its just how things turned out, but theres no guarantee, just like theres no guarantee that going with a simpler design much like the 360 a year after it would have kept kept them in the black, and no guarantee of remaining in this "top spot" of developers especially with all the money that microsoft has, lets not forget that microsoft bit off R&D that sony invested in.....even with all that ps3 still happens to be in the top spot for a lot of devs/pubs, EA especially whos still probably the largest pub when it comes to titles

their loss of billions is applicable to a cost of tech/infrastructure that can be spread out over a single generation and doesnt need to recouped off ps3 and ps3 alone..... things like cell to next gen cell, bd to bdxl, the initial investment of putting together a central network infrastructure such as PSN and its ongoing improvement and expansion into more and more countries..... this so called "disaster" extends beyond cell & rsx, and this "return of the king" mentality is people making it out to be more than it is, its really just business

as for "lazy devs" - i wouldnt use that phrase like a lot of people on the internet do, id just say that some devs have chosen where their priorities are, when your game runs like crap on pc (not the in the way that a game like crysis runs like crap, im talking other games) how much hope is there for ps3 when pc should be the easiest platform out there...... as for devs preferring development on ps3 > 360, im not going to mention any because i dont want to put words into any specific devs mouths but theyre out there, i will just say that im sure most if not all devs will tell you that ps3 dev tools have improved a lot since the start of the gen

on the topic of future porting, why should sony be so concerned with 'generalizing' the architecture for 3rd parties when most studios have already 'solved' the initial porting problems with cell except for those who fall under those two categories mentioned in my previous post which really isnt many at this point in the generation, not to mention pre-ps4 development already taking place on ps3 would port easier to ps4 with cell

- what about devs who actually want sony consoles to shine (1st party) and who continue to develop new techniques that run well on cell along with further optimizing current techniques to run better on cell, why shouldnt sony continue to give these devs more opportunities with an architecture that hasnt been fully explored --- should sony just stick in generalized "cpu" into the ps4 while they wait another 5 years for the gpu space to catch up (ala tile-based deferred rendering)

 

[function]

they didnt plan on losing billions, its just how things turned out, but theres no guarantee, just like theres no guarantee that going with a simpler design much like the 360 a year after it would have kept kept them in the black, and no guarantee of remaining in this "top spot" of developers especially with all the money that microsoft has, lets not forget that microsoft bit off R&D that sony invested in.....even with all that ps3 still happens to be in the top spot for a lot of devs/pubs, EA especially whos still probably the largest pub when it comes to titles

I don't think Sony will take such an easy come, easy go attitude to what's happened with the PS3. It's true that there are no guarantees, but that doesn't mean there's no such thing as a mistake or wrong path. Kutaragi seemed to get canned because of PS3 and/or Cell so Sony weren't happy to let things slide. This doesn't mean that both the chip and the console don't have merits, just that the lack of success was unacceptable from a business POV.

their loss of billions is applicable to a cost of tech/infrastructure that can be spread out over a single generation and doesnt need to recouped off ps3 and ps3 alone..... things like cell to next gen cell, bd to bdxl, the initial investment of putting together a central network infrastructure such as PSN and its ongoing improvement and expansion into more and more countries..... this so called "disaster" extends beyond cell & rsx, and this "return of the king" mentality is people making it out to be more than it is, its really just business

The "something of a disaster" was all the things I mentioned previously, and more. Cell and RSX are symptomatic of focusing on the wrongs and of losing focus. Sony have been working hard to correct things, and PS4 won't be like PS3 (or have all its associated costs).

Cell was a bold move, but history has shown it to not be the right thing for Sony (or Toshiba or IBM) to have focused on.

as for "lazy devs" - i wouldnt use that phrase like a lot of people on the internet do, id just say that some devs have chosen where their priorities are, when your game runs like crap on pc (not the in the way that a game like crysis runs like crap, im talking other games) how much hope is there for ps3 when pc should be the easiest platform out there...... as for devs preferring development on ps3 > 360, im not going to mention any because i dont want to put words into any specific devs mouths but theyre out there, i will just say that im sure most if not all devs will tell you that ps3 dev tools have improved a lot since the start of the gen

In some ways PC development can actually be quite difficult - you have to run (and test) on lots of different hardware, bad drivers or driver changes can screw things up, and from what you hear the PC version gets under resourced compared to the console versions. It always seems to be the PC version that gets squeezed when time runs out ...

I'm sure some devs do prefer development on PS3 to Xbox 360. Some preferred development on Saturn to Playstation! I can't see how development on the PS3 would ever be easier "easier" though. There might be some things the PS3 can do better, or something the 360 (with it's edram) can't do, but the actual process of writing code, debugging and fitting everything in memory would at best be similar I'd have though.

on the topic of future porting, why should sony be so concerned with 'generalizing' the architecture for 3rd parties when most studios have already 'solved' the initial porting problems with cell except for those who fall under those two categories mentioned in my previous post which really isnt many at this point in the generation, not to mention pre-ps4 development already taking place on ps3 would port easier to ps4 with cell

The only thing that will matter for Sony's next gen CPU is handling multiplatform games effortlessly. If cell can do that then fine, but we don't know what MS will be going with. It's also risky to make assumptions about where the next killer app will come from and how much money and talent will be put into your version of the game. There's no need to limit the pool of developers that can make it.

Now that developers are actually up to speed with Cell development, what has the payoff actually been aside from it being an effective crutch for RSX?

- what about devs who actually want sony consoles to shine (1st party) and who continue to develop new techniques that run well on cell along with further optimizing current techniques to run better on cell, why shouldnt sony continue to give these devs more opportunities with an architecture that hasnt been fully explored --- should sony just stick in generalized "cpu" into the ps4 while they wait another 5 years for the gpu space to catch up (ala tile-based deferred rendering)

Consoles shine through the games that they enable, and consoles make money through the games that they sell and that sell them. You don't develop a console to give to give first party developers something expensive to play with.

For first party games it doesn't matter what the hardware is, just so long as it doesn't stink. The last exception to this rule was Halo, ten years ago.