The Future of the Playstation

[one]

However, the strength of the PS3 as a console is and will always remain in the fact that it's a closed system for which software can be optimised to the max. However expensive the hardware may seem at this point in time, this turns it very quickly into a platform where the software (games) make the most of the hardware, and thus becomes very cost-effective, at all sorts of levels. You are, I think, mistaken to think that upgradability is a strength for PC gaming. It is not. The strength of PC gaming, more than anything, lies in the openness of the platform. I could build a PC game right now, no licenecs or expensive dev-kits required. Sure, it is possible to continually upgrade and get more performance out of a PC, but it's going to cost a lot more money in both terms of hardware and software level. And no matter how many PCs are out there, this is what makes PC gaming a relatively small group that is still losing ground to the consoles (it's not shrinking at this moment vis-a-vis itself, but the consoles are still growing faster).

On the other hand, I do think that there are possibilities for creating a game world in which the work of all PS3s involved can create a more immersive world together. It's not going to be easy, and it may require some very novel thinking, but I definitely think there's a lot of potential there.

I don't liken it to fully upgradable PC gaming, instead I mentioned Mac in my 2 posts in this thread. If you don't want to go the full Nintendo way and want to smooth out the transition after PS3 (or realize the 10-year life of PS3 by renewing PS3 itself not to lose competitiveness), the only economically viable option is selling a higher-spec hardware for a higher price. The way PS3 was designed makes PS3 a Cell PC after all, with a fat software layer and reserved RAM. Game developers can get used to a new hardware early by developing an actual game for 32 SPEs, SCE can start mass-producing them without taking a loss while maintaining a certain edge in terms of computing performance unlike Wii. So it's about big cooperation with game developers toward the next generation.

Likewise, at the same time lowering the entry bar for game development by PSN, there'll be more systems of user-created contents or MOD for games as seen in LBP and UT3, along with Java minigames for Home. Basically it's about finding a sweet spot, you don't have to be radically open like the PC market. If its controlled environment can make software publishers feel more secure than the PC market, that's the success. Home is more secure than Second Life, Blu-ray is more secure than DVD, and so on. This way the good part in the PC business can be imported to PS3.

 

[Nesh]

I didnt quite follow what has been discussed so far in the thread but something I wondered it might prove quite useful:

How about making the next PS4 have access from the HDD's memory for some extra MB to be used for either system or video ram requirements?

I think some PC GPUs already do that. Or am I wrong?

I am sure it will save a lot in game development without developers worrying too much about memory and Sony wont have to worry much about increasing production cost by adding extra RAM

 

[Crossbar]

That's one of the reasons why they are touting the 10-year life of PS3. The issues of slower and harder process shrink and a power envelope that is nearing a dangerous point for a CE product. So you can't expect a 10x or more leap anymore for the 5-year cycle. What's possible for the client side is more gradual, scalable updates for an existing format. If you want to maintain the growth pace in terms of entertainment experience, you have to turn your eye to the server side where you can stack up thousands of computers with a smaller latency.
Of course everyone suspects it's why Kutaragi is gone. But the fact is the die is cast and we have the PS3 hardware as it is, left people have to carry on his game to some extent while Hirai handles software license business to recoup the investment.

The intent of this plan is, IMO, to get out of the risk that visits console makers every 5 years. When Microsoft entered the game 7 years ago, it could be easily expected that it'd get harder or impossible to be unharmed by the loss-leading practice of the game console market against the deepest pocket in the world, regardless of winning the market share. Cell and the PS3 hardware format is the last and biggest investment by SCE as the answer to this problem. The plan itself is legitimate and Cell could meet a deadline, but other issues such as software management failures for games, OS and SDK, delays due to Blu-ray and HDMI 1.3, and the economically difficult in-house semiconductor process development after 65nm, and the launch day confusion are more related to why Kutaragi had to leave I think.

How Nintendo coped with it is as you know by using an updated previous-gen hardware with a new controller as a means for product differentiation, and it's very successful for now. PS3 has Cell and Blu-ray for product differentiation, but obviously they both add to the cost and take time to exploit unlike Nintendo's solution. You don't want to compromise on computing processing power either. The only way that makes it possible without being like Nintendo is to sell goods for their fair prices. Or in Mac's or VAIO's case, more than their raw hardware prices with added brand/service value. Sony sells a $200 DVD player and a $2000 DVD player, you can watch the same content on them but in a different way. Why not do it for computer games, why does everyone have to buy a new-gen console at the same time? They know consoles get price drops and PS2 is still growing in emerging markets.

The current PS3 doesn't have enough RAM for PC-like activity. The firmware gets many AV-related updates recently, it's a way to make it a good media player (or a recorder with a tuner), not a computer per se. To run Linux smoothly side-by-side with a realtime OS (Game OS) on a hypervisor like the Toshiba Cell Reference Set, it needs 1GB XDR RAM like the Cell Computing Board. But these added computer capability or superior AV capability sell only to a very small number of enthusiasts even though they can sell this version of PS3 for $1000. The only way to make more people swallow a higher price and establish the game business at the same time is the capability to run a game better.

As for "Crysis scales back 2 years" and game development for a new faster PS3 format, let's see archangelmorph's counterarguments.
Basically my current forecast is based on this comment by Kawanishi (the current CTO at SCEI) that says they don't exclude the possibility of a new PS3 that can run the same game better.
http://forum.beyond3d.com/showpost.php?p=970846&postcount=23
I think there'll be developers who support it, even though they are not first/second-party. PC developers are doing it already, and you know they are complaining that Cell is harder to program and they have difficulties to port 360 games to it, not to mention porting PC games to PS3. PS3.5 makes them easier, you can even port Crysis and whatever to it without losing fidelity to the original PC game. Why Crysis scales back only 2 years is they can't optimize it to a lower-spec PC hardware unlike a fixed console spec. As for segregation, why do they care when Madden runs 30fps on some PS3 and 60fps on the other, as do today on PS3 and 360? OTOH the existence of PS3.5 makes the life of PS3 longer and more competitive. You don't have to argue a 1080p screenshot with AA as a bullshot anymore! As for scalable hardware, as long as they are in the same linear spec hierarchy and don't have more than 3 models, I think it won't be a problem. Recently there was an encouraging comment from one renowned developer - John Carmack. He praised Mac for its uniformity when he was promoting the new engine tech.

To recap my forecast, it'll be like this. In 2010 a PS3 with 32 SPEs and 2GB XDR2 RAM will launch for a price you feel a bit expensive. This time it'll be released before or at the same time Microsoft and Nintendo release their next consoles. PS3 games will begin to support it by running it at higher frame rate, higher AA, more objects, or other bonus as you see in PC games options on different spec PCs. Then, a year or 2 later with enough PS3.5 on the market, there may be a new game profile that requires PS3.5 and doesn't run on PS3 - this is a conservative form of PS4. If the future network infrastructure allows, a new server-side entertainment form with a PS3 as a thin client may appear. Which one is called as PS4 depends on market situation at that time, but they are not mutually exclusive. Also a PS3.5-only game may be introduced at the same time as PS3.5 release to stimulate demand. Anyway my bet is PS4 = PS3.5 + massive Cell server farms.

It's always a pleasure to read your posts, you put a lot thought into your posts.
Edit: I found this reference particulary interesting:

"With more hardware performance, it's possible that the game format also reaches a higher version. Of course the compatibility with the older PS3 is maintained, but on a higher machine you can have a better experience in a scalable way. In other words, the game itself will be the same one though there are different parts depending on hardware performance in the future environment." (says Kawanishi)

Would you care to elaborate on the type of content that you expect be produced by the server farms?

Pachter predicting that a Wii HD will be offered by Nintendo within a couple of years kind of supports your theory that we are heading towards shorter incremental cycles of console hardware.

Do you think that Cells with 8 working SPEs will eventually find their way into Playstation 3? I mean the yield will probably go way up when the 65 nm transistion is completed and I don´t know what kind of device would need 8 SPE Cells in such high volumes, IBM will probably be using their Cells with enhanced DP for their HPCs.

Such an upgraded Playstation 3 could maybe be offered already by Christmas 2008 perhaps with some minor upgrades such as dual HDMI and some more RAM as I earlier suggested?

By christmas 2010 Cell2 may enter the market as you predict together with some other major upgrades of the hardware spec. Though 2010 sounds a bit early in my ears, but Sony may have some bad experience from entering the market late, and as you pointed out the software framework and the global network will be in place right from the start next time which may completely change the pre-conditions for a fast introduction.

 

[weaksauce]

Damn, it's too bad Kutaragi isn't on the team anymore, I think the passion will run out of Playstation without him.

Considering the Cell architechture is the best one can make of transistors, a cell2 wouldn't be bad at all. It would be great for backwards compability and development because of familiar tech. But I wouldn't want to see a Cell3

 

[Shifty Geezer]

Pachter predicting that a Wii HD will be offered by Nintendo within a couple of years kind of supports your theory that we are heading towards shorter incremental cycles of console hardware.

Only if the prediction comes true!

 

[archangelmorph]

To be honest one I don't agree with your predictions at all.. I want to give you some credit since you articulate well and you did put alot of thought into it but I just couldn't see such a scenario ever happening for one main reason..

The Wii..

(let me elaborate..)

To me it seems that this generation started off with the two biggest players reaching for the sky.. They looked over at the PC space to see what was exciting customers (fancy graphics, high resolutions and oversized processors..), took a quick peek at the A/V space to see what was coming over the horizon (HDTV market introduction & HD optical disc storage) and then decided to "put it all in"..
This seemed to go over well with the press and consumer base and it seemed from the offset that the two powerhouse console were set to dominate as their pre-decessors had done prior..

The problem was nintendo decided to be different..

They took a separate approach which emphasized the core function of the box over anything else.. The ability to entertain..
From there, the venture paid off and the Wii exploded onto the scene causing everyone to wake up and realise something very significant..

Not the fact that consumers favour motion sensing over traditional controllers or that games with gimmicks sell.. Not the fact that you can sell a stripped-down underpowered system which cost you peanuts and turn over big profits.. But they introduced the fact that with the right balance of hardware price/performance/features you can not only push out a platform that is profitable for you from the start (and provides a basis for developers to support it without significantly greater cost, making it rather lucrative for everyone, not just the manufacturer..) but with the right angle (or gimmick, or key selling point.. in this case the motion sensing controller..) people will forsake all their desires for the highest fidelity experiences (which cost too much to provide) and jump on board..

This generation nintendo have proved that if your fundamentals are solid (providing a throughly "new" entertainment experiences, not necessarily a fundamentally better one..) you stand a much greater chance of success because whilst, on the surface it may look like consumers are shallow and all they care about are the high end visuals, they actually care more about evolutionary substance in the software which, luckily for everyone is much less expensive to provide..

In light of this I would like to take a step sideways for a moment and give a little insight into my own predictions for the future of PS and the games industry as a whole..

In terms of hardware I've have to say I pretty much agree with Ol' Dennis Dyack to a degree with respect to his vision of a single, "Standardised" hardware specification that all vendors would/could agree on.. To me it would make sense from a financial point of view in so many ways across the board..:-

- Hardware vendors don't have to spend/lose more trying to "differentiate" their hardware platform using more powerful processors under the hood
- R&D costs would be shared and thus reduce the development overheads further
- In a world of ever rising development costs it's pretty obvious that the next round of consoles will be more powerful than the current (to whatever degree..) a unified hardware standard would mean development costs could be reduced by allowing developers to write for the single platform and not have to worry about porting costs, plus having the benefit of the largest possible install base regardless of platform..
- Software vendors could incrementally develop very advanced and sophisticated codebases which would never need to be ported/thrown out for a new platform AND could be shared across the entire industry on the whole which would truley benefit everyone
- Software vendors would then have a setting very similar to the film industry whereby the bulk of [development] investment over time would become more and more content-specific over tech-specific as massive re-usable code libraries could be built up and maintained easily significantly reducing the barrier to entry for new smaller companies who want to create software which can compete with the big budget titles in terms of quality and production values..
- Platforms can "differentiate" in different ways..

Now the last point is one I'd like to elaborate a bit more on..
IMO the Wii proved that the "angle" a console needs to sell isn't exclusively tied to the box's processing power and one could argue just how well nintendo would be doing if they had instead bundled the wiimote with a GC instead..(possibly pretty well but who's to say..) To put it another way, if you took the Wiimote away from the Wii then what *really* do you have? (GC1.5..? Something worse than the Phantom..?)

Now in this world of standardised hardware specifications, individual vendors could provide a similar method of differentiation by creating their own "angle" as nintendo did, whether being a new control mechanism, input device or whatever.. The main point being that the difference would come from the intention to create a new fundamental "entertainment experience" which is unique to the brand, whilst still supporting the more traditional experiences effectively.. This would mean that the focus would be on innovating and stepping away from conventions more and more in order to increase profitability as opposed to adhereing to them consistenly..

I think this could really work out if it ever happened and the benefits would run right into every area of the industry:-

e.g.

Hardware - imagine 3 vendors selling 10s-100s of millions of units in the platform's life-cycle where EVERY single unit turns a profit.?? Do you truely understand how lucrative that could be? (could possibly lead to more 1st party software investment, cheaper MRRPs for games across the board leading the greater software sales, leading to more revenue..)

Software Development - cheaper dev costs, greater focus on making great games with less hardware related investment to worry about as time goes on, leaner hardware learning curve with greater access to shared developer resources, more room for smaller companies to stay a-float and actually turn a profit.. Which would lead to less risk associated with publishing which would lead to greater software diversity and potentially a rise in the mythical "art-house" gaming..

In the end of the day, having console sit at the cutting edge of the hardware curve is always a nice feature, but I say if it can be done cheaply then do it, if not then take a few steps back.. Visual fidelity is fast reaching the point of diminishing returns anyway and therefore it doesn't have to be the priority of the hardware vendor to push for it at the expense of providing a more profitable, better supported and more consumer-friendly-priced box.. Software developers are ALWAYS going to push the envelope with what they can achieve with a given hardware platform anyway and so your still not going to lose out on games not looking better on the platform years after devs have had time to learn more and more of the intricacies of it..

Lastly it really doesn't matter whether doing this will mean console hardware will become obsolete relative to the PC space quicker, since even PCs become obsolete relative to the space in which it resides and this speed of technical advance never really changes the fact that PC dev profitability really only sits in the realm of making games that appeal to the low/medium-end and don't rely on Crysis-like visuals as their "angle" (See WOW, the Sims and the casual games market..)
It would just seem silly to me for console vendors to continue to force some form of consistency between their own models and that of the latest PC hardware since, you only have to look at just how profitable high-end PC is to see that it's not really the wisest thing to do from a financial perspective..

Well that's my predictions anyways..

 

[Color me Dan]

I still don't believe that the next Playstation, not untill broadband internet at 100mbit is as common as dsl or modems are today, will be server dependant. The simple reason is because gamers want to play games whenever they feel like it as long as they have a TV and the box. I know i bring my PS2 with me when i go to a friends cabin, and if i couldn't play the games i wanted to without an internet connection i would never have bought it. Not to mention that some people still don't have internet. Kids rooms might not have an internet jack, or their family wireless.

Look at Microsofts resent misshap with the WGA server; One day you start the console and you can play NO games whatsoever, or the ones that rely on serverbased services. So personally i don't think Sony would go that route unless the technology is there, the market is ALL there with it and you can secure 99% server functionality at all times.

A PS3.5 i think as you describe it One sounds more like an inroad into PC space, and most people get a PS3 to get away from their PC. There is no need to segregate the console market with two or more editions of differently powerful hardware. And even though that hardware might not be used by the games it would still be such a niche market for it that i don't think they could recoup the enormous costs of factory re-engineering on those few machines.

I, like Nesh, wonder if solid state drives could become so fast as to get close to memory and thusly give devs a choice to use it as such if they can hide the delay.

 

[Crossbar]

Only if the prediction comes true!

Very true, that guy predict a lot, and he should as he is paid to do that.

Here is what he said:

Pachter's comments came in response to those made by SEGA executive Scott Steinberg, who expressed concern about the long term future of Nintendo's console.

........

With regard to Steinberg's suggestion that developers are reaching the limits of creative possibility when it comes to the Wii's remote control, Pachter said, "I can't begin to imagine what is on the drawing board for the Wii, but I can say that most developers I've spoken with are extemely excited about the potential for the console... I believe that we've only just scratched the surface.

"Also, Scott's analysis presupposes that we have seen the last in hardware innovation from Nintendo," he went on.

"I disagree. It's easy to envision a Wii 2 in a couple of years that runs at full HD, and has both a Wii-mote and an analog controller, so that all games can be ported to it.

"If Nintendo were to introduce such a device, it would be fully comparable to the Xbox 360 - perhaps it wouldn't have Blu-ray, so a comparison to the PS3 may be unfair - and would likely have most of the same third party content as the other two devices."

But Pachter did agree with Steinberg that the PS3 "will ultimately come out on top". Victory, he predicted, will be the result of a console price cut to USD 199 and the success of Blu-ray.

Price point is key, said Pachter - observing that 80 per cent of all Xboxes sold in the US were purchased for USD 199 or less, with the figure approximately the same for PlayStation 2.

The more I think about it, the more I agree with one that incremental console upgrades is a way of mitigating the financial risks coupled to the quantum leap console generations.

@archangelmorph
Concerning standardized console hardware I can´t see it happening for two main reasons.

1. I can´t see any incenticve for the hardware vendors to go there, what´s in it for them? How can they differentiate their products and make money, not just by having the cheapest manufacturing? Who wants to play by those rules?
2. Who would head the standardization body and how would it be chosen? If the standard is to strict it will limit innovation and if it is to loose it will not provide any benefits for the developers.

It sounds like a dead end road to me as long as consumers are the ones buying consoles. If some joint venture of software publisher started to hand out consoles for free, the scenario might change, but the question that would remain in that scenario is:

Would that be a good console?

 

[Gubbi]

I completely disagree with One that we reached the point of diminishing returns yet. The end of silicon scaling is coming, but there's still a lot of generations left.

When we reach the final proces node there'll still be a lot of technological innovation. Today the investment into equipment for a new node must be amortized over a few years of silicon production. Once we reach the end node, silicon cost will be derived directly from the production cost alone. That will change the economics quite a bit, and we will see substantial more silicon in future products then.

On top of that we'll see new packaging techniques which will allow tighter integration between processing dies and memories.

The second console vendors stop pushing the technological envelope is the second they will die IMO. That is why it is my firm conviction that the Wii will die unceremoniously once the hype dies out. But to be honest, I've only played Wii Sports Bowling, which was a riot. I just cannot see myself sinking 200+ hours into a bowling game the way I do with "normal" console games.

Cheers

 

[Shifty Geezer]

A PS3.5 i think as you describe it One sounds more like an inroad into PC space, and most people get a PS3 to get away from their PC.

The best reference hardware here probably isn't eh Mac but the Amiga. You had the A500 which was in many households a glorified games machine, and the A2000 which could play the same games but functioned as a workstation. A2000 was never intended to be sold as a games machine, but it could serve that purpose. A PS3PC would have the same console hardware and run the same games software, but be intended for applications. If Sony went this route and were smart, they'd enable some improvements to games where possible, such as a 4GHz Cell and 600MHz GPU offering better framerates, and sell to seriously hardcore gamers on top of their workstation market. No dev would ever target this machine specifically though. Any scaling of the game experience would have to be automagical. If Sony were serious about upgrading the PS platform over the life of the machine, PS3 would come ready to be upgraded. Specifically, like the Amiga, at some point Sony would offer a memory upgrade module to 1GB of RAM. Some games would then include bonuses with this RAM upgrade, until the install base was significant enough that some games would require the upgrade.

Without an upgrade path for existing customers, the PS platform is tied to the entry level 3.2GHz, Cell, 500MHz RSX and 512MB RAM.

 

[Arwin]

I was thinking of the Amiga/Atari ST period too. I have a very interesting example in that respect, relating to the ST version of Gauntlett 2. I started out with a 512KB Atari ST with a single sided drive (Atari 520 ST)

- When I upgraded my drive to double sided, I got sampled music in the intro.
- When I upgraded my RAM to 1MB, I got more samples during gameplay.
- When I soldered two Arcade joysticks to the serial connector I got 4 player support (ok, that wasn't quite the same category )

In general, though, I didn't like this kind of thing and I don't really feel it's something that should be supported. The PS3 should be a fixed platform for games.

On the other hand, if you consider the PS3 as a Linux box, then in terms of Linux games I'd be fine with it if games worked like on the PC. But I would still expect the vast majority to target the PS3's fixed base level hardware.

On a more general level, the PS3 has targeted a 10 year lifespan. There are some big advantages in doing so for developers and consumers alike. In a sense, Nintendo on the other hand has basically extended the life of the GameCube hardware platform. The only investment that developers have to make in this respect is coding for the Wiimote. Therefore the Wii doesn't need as long a lifespan, although that also doesn't mean it won't have one of course.

And as is mentioned frequently, they could expand the lifespan a little again by updating the graphics capabilities - although I expect them to make a proper generation jump next time, as they'll be in a good position to do so. Either that, or go software/peripheral/handheld exclusively.

The PS3 on the other hand is supposed to last 10 years, as did the PS1 and PS2. And it well may. The biggest question for me is how far the PS3 will move into the PC and media hub space. That'll be interesting. The first requirement for that would be that the RSX is opened up in Linux. Then shortly after that, we'll need versions that can have upgradeable XDR memory through the super companion chip (as that's now obviously what it is for, witnessing the Cell processing board that was demoed this month)

 

[one]

In terms of hardware I've have to say I pretty much agree with Ol' Dennis Dyack to a degree with respect to his vision of a single, "Standardised" hardware specification that all vendors would/could agree on.. To me it would make sense from a financial point of view in so many ways across the board..:-

That's an odd conclusion after your description of Nintendo's success. I'm pretty sure that such a standard hardware will never happen. Why? Because Nintendo won't agree to it. Nintendo have zero motivation to agree to it because they have had a big success by doing their own thing as demonstrated by Wii and DS. That's why I think rumors such as Nintendo providing their software for iPhone ridiculous. "Lateral thinking of withered technology" (Gunpei YOKOI) is all about Nintendo business and it's the way they create their own standard.

I completely disagree with One that we reached the point of diminishing returns yet. The end of silicon scaling is coming, but there's still a lot of generations left.

When we reach the final proces node there'll still be a lot of technological innovation. Today the investment into equipment for a new node must be amortized over a few years of silicon production. Once we reach the end node, silicon cost will be derived directly from the production cost alone. That will change the economics quite a bit, and we will see substantial more silicon in future products then.

On top of that we'll see new packaging techniques which will allow tighter integration between processing dies and memories.

I hope the growth continues! However do you think such a new process technology can become cheap quickly enough and enable an explosive, fundamental change for people's perception in terms of computer graphics around 2010? Also the LS in Cell's SPE was criticized a lot like selling a soul to devils, but it was to squeeze out realtime performance from the same die size. It's like cheating in a way depending on how you look at it. Do you expect more radical change than this in future? At SIGGRAPH 2007 IBM did a realtime ray-tracing demo, but it required 112 SPEs. I doubt you can cram them into one console in 2010.

 

[one]

Would you care to elaborate on the type of content that you expect be produced by the server farms?

I have no more concrete image than Kutaragi's "you can jack into servers of future Sony Pictures studio where a film director is shooting a movie in realtime!", but there are signs of what to come in a patent discussed a year ago

http://appft1.uspto.gov/netacgi/nph...52".PGNR.&OS=DN/20050238052&RS=DN/20050238052
20050238052 Method and apparatus for providing an interconnection network function
Inventors: Yamazaki, Takeshi; (Tokyo, JP) Assignee Name and Adress: Sony Computer Entertainment Inc. Tokyo JP

[0007] The simulation of the actual world to a user of a processing system as a "virtual world" is another LSCP which requires a high degree of collaboration between processors and real-time sharing between them to provide real-time services. In particular, a high degree of collaboration and real-time sharing are required to provide a virtual world which simulates the actual world and actual sensory experiences from locations around the world, while providing interactive play. In order to make experiences believable to the user, much sensory data needs to be collected in real-time from actual world sites, and "recreated" when the user "visits" the corresponding virtual site in the virtual world. Thus, data representing experiences such as sights (e.g. current images of the actual world site), current sounds, and indications of temperature, humidity, presence of wind, and even smells must be collected and made available to the user.

[0008] Because of lack of a processor network capable of supporting it, such virtual world is an unfulfilled need. It is estimated that the processing requirements for such virtual world would exceed the capabilities of the fastest supercomputer in the world, which is currently the "Earth Simulator", a supercomputer in Japan having a speed of 82 Teraflops/sec, and a latency of 10 .mu.s. The Earth Simulator is believed to be incapable of supporting such virtual world because of high latency, among others. High latency can be caused by high protocol overhead in messaging between processors. Thus, a need exists to provide a network of processors which communicate via a low overhead communication protocol having reduced latency, so as to permit increased collaboration between processors and improved sharing of information in real-time.

and a research like this

Cell Based Servers for Next-Gen Games
http://sti.cc.gatech.edu/Slides/DAmora-070619.pdf

No dev would ever target this machine specifically though.

Actually I think PSN is a good indication of who targets such a machine. Big developers can't move quick enough, but small devs can make a tech demo into a small game and make it profitable. If SCE is really willing, they can make future hardwares more open.

 

[Color me Dan]

The PS3 will most likely be kept competative for 10 years with optional input devices for games and as a general but extended media hub. We're seeing this allready and i'm sure it will be improved upon further. I think Sony has been looking at the Wii alot and they are probably developing controller ideas and talking to devs to sort of bring that audience over to the PS3. By doing so they would also (most likely) have a wider range of input devices which makes for a more varied gaming experiance. In the long run i'm sure Ms will try something similer to keep their console alive and well for as long as possible.

Knowing Nintendo and their "profitable-out-the-door" ideology i think hardware wise such a console would be in range of the current Sony and Ms offerings so that would reduce incentive to upgrade to the next gen Playstation before a 10 year period as well. But as i said, the PS3 will probably get by with controller inovation, just a hunch i have.

This begs the question, how will the next generation of input devices for the Playstation look like? Beyond PS3 obviously.

 

[Gubbi]

I hope the growth continues! However do you think such a new process technology can become cheap quickly enough and enable an explosive, fundamental change for people's perception in terms of computer graphics around 2010? Also the LS in Cell's SPE was criticized a lot like selling a soul to devils, but it was to squeeze out realtime performance from the same die size. It's like cheating in a way depending on how you look at it. Do you expect more radical change than this in future? At SIGGRAPH 2007 IBM did a realtime ray-tracing demo, but it required 112 SPEs. I doubt you can cram them into one console in 2010.

We'll be at the 32nm proces node for the next console generation (in 2011-12). That's roughly an order of magnitude more logic for the same die size compared to 90nm. That's 10x more RAM, 10x more cores, 10x more shaders etc. Will that be evident in games? Yes.

Scaling past the next generation looks shady though because of the limits imposed by Mother Nature. That's when we'll see a change in the economics regarding LSI production which will lead to a change in how it's used. So we'll see a radical change for the next next generation (PS5) too. After that, - who knows ?

Cheers

 

[archangelmorph]

@archangelmorph
Concerning standardized console hardware I can´t see it happening for two main reasons.

1. I can´t see any incenticve for the hardware vendors to go there, what´s in it for them? How can they differentiate their products and make money, not just by having the cheapest manufacturing? Who wants to play by those rules?

1. 
   I've already provided some examples of how they could do this in my previous post..
   
   

   [*]Who would head the standardization body and how would it be chosen? If the standard is to strict it will limit innovation and if it is to loose it will not provide any benefits for the developers.

Thats easy.. All three hardware vendors would agree on the standard.. they themselves would setup the collaborative standardization body and representatives from all three companies (in the form of R&D guys would do the leg work to establish the standard..
Serious this could work really well, especially with respect to the great strenghs each company will be pringing to the table..:-

MS - developer mindedness (providing solutions to ease the dev process) and excellent OS and software services contributions..

Sony - manufacturing oomph, technological-mindedness and designing powerful hardware to provide developers with enough power to do anything they may wish to..

Nintendo - consumer-mindedness, innovative ideas to provide something thats fresh and ideal for everyone..

Sure there's definitely going to be alot of contention due to conflicting visions as the companies stand today but who knows what those visons will be in 3-5 years from now..

It sounds like a dead end road to me as long as consumers are the ones buying consoles. If some joint venture of software publisher started to handle out consoles for free, the scenario might change, but the question that would remain in that scenario is:
Would that be a good console?

I don't understand what your saying.. Why would they hand out consoles for free?

I think alot of people are confusing my idea slightly though..

I'm not saying have a joint venture where all three companies can develop one console for the masses..

I'm saying have one single unified "hardware specification" from which all hardware vendors would derrive their own console designs which have the core spec suitable for games and then anything else they may wish to throw in the box with regards to unique or novel services, connections, innovative peripherals etc..

The benefits being as i've described in my previous post..

That's an odd conclusion after your description of Nintendo's success. I'm pretty sure that such a standard hardware will never happen. Why? Because Nintendo won't agree to it. Nintendo have zero motivation to agree to it because they have had a big success by doing their own thing as demonstrated by Wii and DS. That's why I think rumors such as Nintendo providing their software for iPhone ridiculous. "Lateral thinking of withered technology" (Gunpei YOKOI) is all about Nintendo business and it's the way they create their own standard.

From the Wii's success which was not based on the hardware spec at all but on a peripheral bundled with the machine and games tailored to support it, it's pretty clear that nintendo would have just as much to gain from this as would anyone else..

They would have the benefit of M$ & Sony's expertise in terms of providing a spec that wouldn't have to be underpowered like the Wii (that being the consoles only real limitation on the market currently, however relevant..) and they could benefit from this synergy in other ways such as very rapid enhancements to their in-house codebases from the shared developer resources network (especially where we have so many Sony & M$-related developers who just love this kind of work..) It would make their own development processes more cost effective since they could continue to concentrate soley on gameplay (as would every developer..) & never have to worry about spending money on software R&D to make their games visually competitive since the code would be readily available across the board..

I know many of you don't agree with me on this but I just don't see how this idea of even more PC-like incremental improvements could ever benefit developers as is..

I look around (especially here in the UK) and there's so many fewer developers than there used to be with so many more losing finanial viability due to rising dev costs, high consumer expectations and highly seggragated markets.. I don't think adding to the problem is ever going to make things get better and unless something is done then I'm not sure there's going to be much of an industry left outside of possibly Wii-centric & casual (cashcow) games & the high end (multi-million-dollar GeOW, Halo 3 etc..)..

 

[Crossbar]

I've already provided some examples of how they could do this in my previous post..

Thats easy.. All three hardware vendors would agree on the standard.. they themselves would setup the collaborative standardization body and representatives from all three companies (in the form of R&D guys would do the leg work to establish the standard..
Serious this could work really well, especially with respect to the great strenghs each company will be pringing to the table..:-

MS - developer mindedness (providing solutions to ease the dev process) and excellent OS and software services contributions..

Sony - manufacturing oomph, technological-mindedness and designing powerful hardware to provide developers with enough power to do anything they may wish to..

Nintendo - consumer-mindedness, innovative ideas to provide something thats fresh and ideal for everyone..

Sure there's definitely going to be alot of contention due to conflicting visions as the companies stand today but who knows what those visons will be in 3-5 years from now..

What you are describing is a monopoly setup, people tried it a lot in the Soviet union, it didn´t work out very well for the consumers nor the companies.

I don't understand what your saying.. Why would they hand out consoles for free?

Because usually standards (beside national standards) are created by some unified buyers who setup some requirements (such as a communication spec or whatever) for an item of interest and present it to the manufacturers, the manufacturers that comply to the spec will get the orders.

Well the console buyers are average Joes who buy whatever console that they think gives best value for the money, they will never unite to create a spec and present to the console manufacturers.

The only ones that may benefit from a standard are the software publishers who would get a less fragmented market, they could unite but how could they put pressure on the console manufacturers? By stopping developing games? Not very likely as long as the manufacturers have their own studios (that may be a reason why Sony has invested a lot in their own studios), creating their own console spec and invite tenders and actually buying consoles according to that spec might actually work and they could either hand them out for free or selling them.

But seriously this will not happen.

I think the Software publishers that are concerned about the fragemented market should invest money in middleware that will ease multiplatform development and act as Esperanto for the developers. I think that is a better way to go than fighting for a unified console spec.

 

[archangelmorph]

I think the Software publishers that are concerned about the fragemented market should invest money in middleware that will ease multiplatform development and act as Esperanto for the developers. I think that is a better way to go than fighting for a unified console spec.

Sure it's better for publishers.. but what about developers?

At the moment publishers recognise the costs associated with middleware (see UE3.0) licenses which currently maybe a viable solution for a single-team, single-project looking to release cross-platform but this solution is still far too expensive for publishers to buy and distribute to all developers they contract.. Then there's developing their own middleware which works out even more expensive..

Basically as a result of these high development costs you have fat publishers too scared to invest in projects other than those which either have already been heavily invested in (by the developer, if they have the cash..) to establish a proof of concept (with much of the technology in place) so it's financial and marketable viability can be assessed, Or are established franchises or licensed IP..

Most publishers won't even touch you if your a small scale development house with a big vision and no history (i.e. hit titles under your belt..) and with the costs of development in terms of staying competitve being as high as they are nowadays you're either forced to do outsourcing work (if you can find it) or target a smaller niche (e.g. handheld..) where you can get yourt products out the door cheaply and quickly (albiet with just as small returns..)

I'm not saying a unified hardware specification standard would be particularly simple and straightforward to see introduction in the near future, but in the end something needs to be done to help the current situation change for the better and encourage an industry that is far more lucrative across the board & far less volatile than the one we have today..

If content is the heart and sould of the industry then surely content developers need our lives making easier and not more and more difficult by continuing to increase technical complexity (iterating every few years) and reducing/seggregating financial viability..

 

[Shifty Geezer]

Actually I think PSN is a good indication of who targets such a machine. Big developers can't move quick enough, but small devs can make a tech demo into a small game and make it profitable. If SCE is really willing, they can make future hardwares more open.

How many people are going to buy a more expensive PS3+ for a few niche titles? Surely the whole point of a PS3+ is to enable a better experience in Madden, R:FoM3, and the mainstream titles hardcore gamers buy hardware for? Not to mention the fact that to use more resources on the whole requires more expenditure. PSN titles are mostly smaller titles that don't use PS3 completely. It's hardly a deployment service for the most taxing software with the most assets. For that, with huge amounts of data, you have discs that are more viable. 25GB downloads aren't going to be popular and mainstream for a while yet!

 

[Shifty Geezer]

What you are describing is a monopoly setup, people tried it a lot in the Soviet union, it didn´t work out very well for the consumers nor the companies.

I think the idea is more like VHS or DVD. A standard that everyone develops for. Like PC hardware even, and that was a direction MS were talking about. Future XB's wouldn't necessarily be a fixed hardware platform, but a hardware standard. If anyone's going to go that route, it'd be MS who want to tie the two sides of the industry together. Well, three sides really, with PC, Media PC and games console.