Predict: The Next Generation Console Tech

[Prophecy2k]

It's not that I think the RSX was a poor design choice, it's that TMK, Nvidia is pretty damned hard-headed about pricing for the parts they provide, and the last thing anybody needs or wants to see again is an over-priced box clinging to store shelves. It's bad for business.

The CPU as I see it right now is anybody's guess. I *do* think the CELL still has some life to it, but the field has changed since CELL was just a concept; today everybody has a multi-core CPU, GPUs have encroached, conquered and left, too. As it stands, PS4 could have CELL2.0, some Intel design (they're hard-headed on price, too, but I think they'd cave to leverage Larrabee), they may go for yet another custom IBM job, or something from AMD. We have a lot of options this coming round vs last, and I think it's going to come down to finding the sweet spot of price, long-term proffitability, performance and efficiency.

I think we could easily hit 10x performance, regardless. Another year from now we'll be looking at new video cards that *on their own* are much more powerful than that, and by the time the next generation of consoles actually hits, those parts will be budget bits by comparison. If the next generation starts in late 2012, and we actually get stuck with stock-ish GPUs and CPUs from late 2011, we'd still have something nearly twice as fast as a theoretical Radeon 6xxx in there, and any number of CPUs that beat the snot out of a Core i9, CELL, or any current AMD part. And that I feel could represent a budget-friendly box.

The thought of this does make me rather giddy indeed

If Sony do opt for a custom ATI design and microsoft do too, I wonder how similar both parts could end up being in the long run?

Also, if Sony decide (I hope they don't) to kill CELL and opt for a different CPU, what would you think would be most likely taking into account issues like backwards compatibility and such?

 

[liolio]

In regard games size I'm pretty much with Barbarian. I wouldn't expect games to go past the 25GB for many reasons the most relevant being costs and business model.
* BRD are likely to be standard next gen, I can see editors and publishers choosing the cheapest optical format out there which will be BRD single layer.
* Editors publishers and manufacturers will try to tackle the second hand market. Online sales is clearly part of this equation. Keeping games size "minimal" will help here.
*DLC importance in the business will grow imho. I expect the game industry to move to more episodic content. Rockstar and MMORPG providers are ahead of the overall industry in this regard. Some costumer may dislike it but there are a lot of benefits for the industry to do so. Overall it diminishes the need for a big "first issue" for a title and thus the need for high capacity optical format.
*Live arcade/psn games size is likely to grow, see the point above
*I can't see manufacturers going for huge and too costly HDD, actually I can see them going with the cheapest options.

Overall I don't think that ID tech can be use as an argument to claim that we will need xxGB optical format, no matter the merits of the tech if it's a bad match for the industry business model it will go nowhere (or will have to evolve).
Some games could use two layers BRD (I expect manufacturers to up the royalties in that case to keep their number minimal).

 

[ShadowRunner]

It's not that I think the RSX was a poor design choice, it's that TMK, Nvidia is pretty damned hard-headed about pricing for the parts they provide, and the last thing anybody needs or wants to see again is an over-priced box clinging to store shelves. It's bad for business.

It was my understanding that Sony own the IP for RSX, i didnt think Nvidia had any input on the price of the part. They would have been contracted and paid upfront to design the GPU and from there it was in Sonys hands. That was my understanding anyway could well be wrong...

 

[Prophecy2k]

In regard games size I'm pretty much with Barbarian. I wouldn't expect games to go past the 25GB for many reasons the most relevant being costs and business model.
* BRD are likely to be standard next gen, I can see editors and publishers choosing the cheapest optical format out there which will be BRD single layer.
* Editors publishers and manufacturers will try to tackle the second hand market. Online sales is clearly part of this equation. Keeping games size "minimal" will help here.

I think this is an important factor as I still don't see global consumer internet connection speeds improving enough by 2013 to allow any significant chunk of the market the ability to download a 25GB game in a reasonable timeframe.

In that I wouldn't expect the same 25GB retail releases to be made available for DD, it'd just be too much bandwidth for PSN and XBLA. The complete adoption of single layer BD next-gen would make it even harder for publishers to combat retail used game sales with DD, hence why I don't think it would really help publishers in their efforts to conqueor the used game market.

I'm not sure about your point on episodic content though too, since whilst this may work for MMORPGs (of which there are currently next to none on consoles) and GTA (i.e. relatively quick to create scenarios in the same open world environment), I'm really not sure how well episodic content will work with other games and genres. It certainly hasn't faired too well with the half-life series (in terms of dev time and release frequency).

Again I'm not sure that the global gaming market is really all too keen on episodic content if only for the value proposition aspect. How many DD copies of GTA IV: TLATD were sold in comparison to retail copies of GTA IV on the 360? The sales success of such products will ultimately dictate where publishers go next.

I'd also be bold and suggest that recent trends show that publishers are wanting to add a lot more to your $60 retail package (with online multiplayer options alongside a sizeable SP campaign) than dropping the price and shipping smaller retail games at a lower price point.

Yes I agree that episodic content will likely become more successful with the availability of DD content, however as of now retail titles sell much more and generate far greater profits for publishers (albeit with the added risk and exposure to the rental/used game markets to steal a good piece of your pie).

I don't think episodic content will play a significant roll yet for maybe one or two more generations, unless someone comes up with a radically different business model for improved content release frequency and game design (Interesting question: how long can R* keep churning out GTA IV episodes before people get bored of liberty city?).

 

[liolio]

I think this is an important factor as I still don't see global consumer internet connection speeds improving enough by 2013 to allow any significant chunk of the market the ability to download a 25GB game in a reasonable timeframe.

In that I wouldn't expect the same 25GB retail releases to be made available for DD,it'd just be too much bandwidth for PSN and XBLA. The complete adoption of single layer BD next-gen would make it even harder for publishers to combat retail used game sales with DD, hence why I don't think it would really help publishers in their efforts to conqueor the used game market.

I think it's more a storage issue than bandwidth and they will have to swallow the bullet anway. But I disagree I don't have a crazy internet connection and if the games available through game on demand were not that old, I would go for it no matter the time it takes to download. In fact xbox live is surprisingly fast even on my not that good internet access. For those who can wait some hours or a night there is still BRD. Being able to play without swapping disk is worse it. I would not be surprised if Ms reduces the gap in the future between physical and on line releases. They have to find a equilibrium without losing support from retailers.For those who can't wait some hours or a night there or are interest in the advantages of physical format well BRD will still be there.

I'm not sure about your point on episodic content though too, since whilst this may work for MMORPGs (of which there are currently next to none on consoles) and GTA (i.e. relatively quick to create scenarios in the same open world environment), I'm really not sure how well episodic content will work with other games and genres. It certainly hasn't faired too well with the half-life series (in terms of dev time and release frequency)

Again I'm not sure that the global gaming market is really all too keen on episodic content if only for the value proposition aspect. How many DD copies of GTA IV: TLATD were sold in comparison to retail copies of GTA IV on the 360? The sales success of such products will ultimately dictate where publishers go next.

I'd also be bold and suggest that recent trends show that publishers are wanting to add a lot more to your $60 retail package (with online multiplayer options alongside a sizeable SP campaign) than dropping the price and shipping smaller retail games at a lower price point.

Yes I agree that episodic content will likely become more successful with the availability of DD content, however as of now retail titles sell much more and generate far greater profits for publishers (albeit with the added risk and exposure to the rental/used game markets to steal a good piece of your pie).

I don't think episodic content will play a significant roll yet for maybe one or two more generations, unless someone comes up with a radically different and novel concept for content frequent content delivery and game design (i.e. how long can R* keep churning out GTA IV episodes before people get bored of liberty city?).

It's my personal opinion but I think that downloadable/episodic content that easily extendible to most genres. It's already there for multi player part of most games the next jump is solo/single player experience.
It's true that costumers are iffy on DLC/episodic content because they more than often feel ripped, thinking I'm paying for something that could have been on the game I bought to begin with (capcom behaviours with RE5 is not here to reassure them...). So I guess there is a need for a proper marketing effort about how you market a game split in episode and how you price it too. We don't know how ODST will fare but by the reviews it sounds like Ms may not have found the proper balance for example (sales will tell).
For Rockstar I'm not sure that my memory is right but I remember ~2millions figures and the DLC will be release on DVD too, episodic content doesn't forbid physical release. Rockstar will imho have made the most of their efforts with GTA4 but I've the feeling like they must feel like they didn't charge enough for the DLC (not like they were free anyway). Physical release is a good back up if online sales fails to meet expectations but I feel like it must also allow for cleverer game budgeting. What I mean the first iteration of the "game X1" costs and would have cost XX millions dollars. You know you will release "game X2", it would have take one year to make it. Basically your business plan is to split the part 2 in two parts. You ship the first part only six months later and start getting returns on you investment. It's a lot less risky now to found "game X2" part 2. You ship "game X2 part 2" on line and then (add one or two months) the "full game X2" on BRD.
I see quiet some advantages here for the editors. First they earn money during the process of creating "game X2". They keep the game "hot" and that's something editors for heavily multiplayers oriented games have already understand, players don't want to wait X years between two release of a franchise, they need to get fed content to continue playing. Not every franchise is GTA, GT or halo and is likely to sell well no matter when it's release and won't fall in oblivion in the mean time.
In regard to the second hand market, well it's about the timing of the release and creating desire. You can sold you game and wait for "game X2" release but If you don't you will access content way earlier than you would if you have wait for the physical release. It's even more relevant for games that have a consistent story line, for something like GTA4 you don't really in the urge to know "what comes next?".
I think it's a nice way to capture the enthusiastic part of the market and to tone down their "second hand market habit impulse. You also dismiss the loss due to the "rental offering" for this part of the market.

In regard to the content creation I think that it would not be that tough as long as you know more or less where you're heading with the story and on how many iterations you intend to split the content. I could even see optimization in the utilisation of your existing asset.
In some games you go through different environments almost rush through to the story is a bit cut may have profit from more development, etc. In the end game designers/developers have to make you through a whole story with only finite resources whether it's human time or disk space, I think that removing the pressure to finish the story in one take may ease things quiet a bit and prevent for too obvious dash work.

Overall I hope they will go that route SP campaigns get shorter and shorter, it looks like editors and publisher have realize that multiplayer is right now is a better way to keep players stuck to your franchise (and spending money on your franchise) my trust is that episodic content and DLC could help solo single player gaming to regain relevance.
They have to learn from the Hollywood series business. Costs are going really high when something works they have to make the most out of it till it works. What they need is proper Story and good scenarists.

But I've a bad feeling about that as multiplayer will more than often ending being a cheaper solution... I'm unsure publisher will search for way to fond more consistent SP experience

 

[AlNom]

It was my understanding that Sony own the IP for RSX, i didnt think Nvidia had any input on the price of the part.

Considering that it's almost* entirely a G71, I would think not...

*There are other modifications for XDR access.

 

[ShadowRunner]

Considering that it's almost* entirely a G71, I would think not...

*There are other modifications for XDR access.

True but if Sony contracted Nvidia to produce a chip that it would own IP for and a modified G71 is what they came up with i dont think it would stop it being sonys IP legally. MS owns the IP for its GPU i believe, that didnt stop ATI using much of the same tech in thier desktop GPUs right?

Sony manufactures RSX themselves, this is where the cost fluctuates, and nvidia has no control over this. Sony would pay a liscensing fee but this would be paid up front or would be a fixed fee agreed in advance, i dont see any way sony could be unexpectedly burned by nvidia being hard-headed on reducing cost.

 

[AlNom]

Xenos is much more different under the hood when you consider the shader array setup and how it works in with the rest of the design with the eDRAM. Individual components might be similar, but the entire design is very different from any PC part. It's that design that constitutes a new IP.

RSX is largely unchanged from G7x with only the addition of the XDR/Cell hooks. Disabled ROPs, extra texture cache, or a smaller bus width do not constitute a significant design change to warrant classification as a new IP.

Whether or not Sony got burned is not what I am discussing btw.

 

[DuckThor Evil]

At the same time, digital distribution of games will not be feasible either for the vast majority of customers, not for many, many years. Even the largest laptop harddrive can't hold more than a handful of blu-ray titles, and most gamers don't have the kind of internet connection needed to download dozens of gigabytes in any kind of reasonable timespan.

I think piracy has proved that digital distribution is viable for fairly large number of people. Slower internet connections wouldn't be such a problem if you can preload content before the release date and just download an activation key on the launch day. I think physical media will still remain and it's definitely possible or even likely that MS will choose Blu-ray, but DVD with full or partial installs for multi disc games might happen also.

 

[ShadowRunner]

Xenos is much more different under the hood when you consider the shader array setup and how it works in with the rest of the design with the eDRAM. Individual components might be similar, but the entire design is very different from any PC part. It's that design that constitutes a new IP.

RSX is largely unchanged from G7x with only the addition of the XDR/Cell hooks. Disabled ROPs, extra texture cache, or a smaller bus width do not constitute a significant design change to warrant classification as a new IP.

Whether or not Sony got burned is not what I am discussing btw.

But is there a legal distinction on what constitutes new ip, or is it just your opinion on what should be classed as a seperate ip? Surely it is up to nvidia themselves to decide?

Im pretty certain sony makes it themeselves and pays a licsense to nvidia for using thier tech. I guess they wouldnt have to own the ip do do this so yea you are most likely right. I think MS have the same deal though, maybe they dont own the ip either an licsence it also.

Either way, my primary point was nvidia in no way have direct control of what RSX costs sony to produce.

EDIT: From wiki:

"December 2004 saw the announcement that NVIDIA would assist Sony with the design of the graphics processor (RSX) in the PlayStation 3 game console. In March 2006 it emerged that Nvidia would deliver RSX to Sony as an IP-core, and that Sony alone would be responsible for manufacturing the RSX. Under the agreement, Nvidia will provide ongoing support to port the RSX to Sony's fabs of choice (Sony and Toshiba), as well as die shrinks to 65 nm. This is a departure from Nvidia's business arrangement with Microsoft, in which Nvidia managed production and delivery of the Xbox GPU through Nvidia's usual third-party foundry contracts. Meanwhile, Microsoft has chosen to license a design by ATI and make their own manufacturing arrangements for Xbox 360's graphics hardware, as has Nintendo for their Wii console to succeed the ATI-based GameCube."

 

[sunscar]

There is another reason I don't entirely see Nvidia landing a console license. Their love for monolithic designs. I still think if anything the next generation *for the most part* will have a pretty hard focus on efficiency over brute theoretical power.

 

[Rangers]

Maybe Sony "owns" the RSX IP but cant really do anything with it but fab it and put in PS3, legally. I imagine it's the same for Xenos.

In that case they "own" the IP but it would be so narrowly restricted it doesn't matter.

Although if such agreements exist, I dont know why Intel balked at allowing their IP to be "owned" in such a way by console makers.

Looking at Xenos, obviously MS doesn't want to make a GPU of their own. All they really want is the right to fab it. It seems a legal agreement could thus be structured.

 

[brain_stew]

Maybe Sony "owns" the RSX IP but cant really do anything with it but fab it and put in PS3, legally. I imagine it's the same for Xenos.

In that case they "own" the IP but it would be so narrowly restricted it doesn't matter.

Although if such agreements exist, I dont know why Intel balked at allowing their IP to be "owned" in such a way by console makers.

Looking at Xenos, obviously MS doesn't want to make a GPU of their own. All they really want is the right to fab it. It seems a legal agreement could thus be structured.

Well the fact that Intel own a hell of a lot of fabs might have something to do with that. I'm sure Intel wouldn't want their chips being fabbed by Global Foundries or TSMC, though their eagerness to get LRB into next generation consoles may mean they'll have to be more leniant.

 

[marcus_rocks]

http://translate.google.it/translat...ted-2-a-milano.php&sl=it&tl=en&hl=it&ie=UTF-8

So only a filter was applied, I mean I'm having a hard time believing it myself too so you tell me dude.
The crysis rendering method is confusing to me though, what kind of effect would require such a drastic increase in base model poly count? Wouldn't that cause major visual discrepancy when played?

In the same interview:

The Naughty Dog Engine 3.0 graphics engine is already in production, the team dedicated to this issue is already laying the groundwork for the future

Oh my ..., with this steady progress, do we really need the PS4?

 

[liolio]

Oh my ..., with this steady progress, do we really need the PS4?

Hum, I wonder at first if it was a joke but anyway. My answer is yes because I want way better than that. A more educated answer would come down to how complex next engines will be.
I advice you to read (if you don't have ) end of the gpu roadmap from Epic (available in my sig) or Crytech presentation about what is coming next.

Both plan to deal with massively multi-core systems offering a lot of flexibility and thus offering opportunity for various rendering modes. One thing is constant they plan their next engine to take quiet some time to develop. Rein was wrong and directx 10 wasn't the last relevent API but by watching at where Nvidia and INtel are heading (ATI should follow next but they just don't intensive to do it in a anticipated manner) directx 11 could very well be the last one.
Engine and middleware providers have some naughty challenges in fornt of us to harness the power of next generation systems.
Shortly you should not read to much in Naughty Dogs statement, they have started the development now because it will be a long one not because of an hypothetical ps4 anticipated launch

 

[holsty101]

I thought that reference to ND 3.0 engine was regarding their next ps3 game, rather than for the next hardware generation?

 

[Berek]

I'm particularly interested if the "PS4" and "x720" will have DX11 graphics... or higher. We can be fairly certain they will have dual, maybe even quad core CPUs, finally plenty of memory, and a robust GPU in general. I see the next generation consoles really making a difference in the diversity of games available on them more than ever in previous generations. Unless of course they go the way of the Wii 1 and just recycle existing Gamecube technology in a new package, making tweaks here and there. Then perhaps not this next generation.

 

[AlNom]

You can bet on Microsoft putting in hardware that supports their latest DX version and perhaps prototype functionalities.

 

[Kaotik]

I'm particularly interested if the "PS4" and "x720" will have DX11 graphics... or higher. We can be fairly certain they will have dual, maybe even quad core CPUs, finally plenty of memory, and a robust GPU in general. I see the next generation consoles really making a difference in the diversity of games available on them more than ever in previous generations. Unless of course they go the way of the Wii 1 and just recycle existing Gamecube technology in a new package, making tweaks here and there. Then perhaps not this next generation.

You do realize that XB360 already has triple core CPU, right?

 

[semitope]

I think piracy has proved that digital distribution is viable for fairly large number of people. Slower internet connections wouldn't be such a problem if you can preload content before the release date and just download an activation key on the launch day. I think physical media will still remain and it's definitely possible or even likely that MS will choose Blu-ray, but DVD with full or partial installs for multi disc games might happen also.

The option to install and run a disc based game off the HDD is better. Piracy of games is still pretty much limited to less than 7GB (360 games) so when you get to ps3 games and next gen games its not so easily demonstrated that digital distribution is workable.