Predict: The Next Generation Console Tech

[upnorthsox]

Nintendo does have quite a few good options between Bobcat, Bulldozer and Llano and any various custom chips based upon each depending on how much they want to invest, spend and their target bill of materials.

They can certainly get away with (alot) less than cutting edge and still deliver a significant upgrade. There goal should be to get to a rough content object parity so devs don't need separate content bases like this gen, things like advanced lighting, AA, OMB, etc... can all be watered down if need be.

 

[TheChefO]

Were any people besides diehard nintendo fans playing gamecube games on the Wii? Would it have even really mattered to the fanboys if it weren't there? In practice, this gen was the anti-BC generation for al the consoles. Next gen may be a bit different with all the DD content, or not, still hard to say but since Ninty hasn't really participated in that this gen it doesn't realy matter to them either way,

Nintendo has offered quite a few download games.

And they've built quite a userbase.

The tech gap between wii and current level hardware is large enough where they shouldn't have a problem emulating it so I don't see why they would choose not to.

 

[TheChefO]

I was under the impression BC was as good as dead for a couple of reasons.

1. They can sell recompiled DC versions of the old games. (ie Halo CE for 10 or 20$) and
2. By sidestepping the issue they remove all kinds of legacy problems with software and/or hardware.

The legacy sales made sense this gen, as many people were migrating from one console which dominated (ps2) and are also currently building digital libraries which weren't possible last gen.

To expect users to dump that library and jump onto that bandwagon again nextgen is foolish. Especially if their existing architecture is reasonably scalable or able to be emulated.


In fact, some hardware venders have been forward thinking WRT software development in preparing for BC by keeping developers writing abstract instead of to the metal.

 

[upnorthsox]

Nintendo has offered quite a few download games.

And they've built quite a userbase.

The tech gap between wii and current level hardware is large enough where they shouldn't have a problem emulating it so I don't see why they would choose not to.

Nintendo regurgitated a bunch of SNES type games, those are being played on smart phones now. Next gen they don't want you playing Super Paper Mario, they want you playing Super Paper Mario 2DHD!!!

 

[upnorthsox]

The legacy sales made sense this gen, as many people were migrating from one console which dominated (ps2) and are also currently building digital libraries which weren't possible last gen.

To expect users to dump that library and jump onto that bandwagon again nextgen is foolish. Especially if their existing architecture is reasonably scalable or able to be emulated.


In fact, some hardware venders have been forward thinking WRT software development in preparing for BC by keeping developers writing abstract instead of to the metal.

Well we'll see how that goes I guess. They gave it alot of lip service at the beginning of this gen and then stuck a stake in it faster than Van Helsing,

The DD is a factor though.

 

[brain_stew]

I was under the impression BC was as good as dead for a couple of reasons.

1. They can sell recompiled DC versions of the old games. (ie Halo CE for 10 or 20$) and
2. By sidestepping the issue they remove all kinds of legacy problems with software and/or hardware.

Hey, I could be dead wrong. However, it would not surprise me if only a few big name titles actually still worked and they allowed publishers/ developers to do a code rework and sell the games all over again, at a much reduced price of course.

Sidenote:
Are we really talking about 2013 here for everyone? That seems like an incredibly long time between generations. That would be 8 years for the 360!

Late 2011 for Nintendo.

Late 2013 for Sony and Microsoft.

Would Nintendo really need to include dedicated hardware in order to make BC work on the sort of hardware we're suggesting? The gap between the Wii and its successor is going to be huge in terms of raw power no matter what we end up with and AMD can always make slight tweaks to the GPU hardware in order to make BC less of a challenge (although I doubt even this will be necessary considering similar PC GPUs are managing to render Wii games in HD with supersampling just fine). Heck, is sticking the Wii's PowerPC core on die really going to affect costs all that much? That thing is going to be tiny at 28nm and not have any real noticeable effect on die size, but again I doubt dedicated hardware like that is even needed to provide an adequate BC solution. Software emulation will more than suffice and its an area that Nintendo already have a big investment in. I'm sure they'll already have an internal x86 based Wii emulator, it'll just be a case of optimising it and ensuring compatibility with most major releases. It doesn't have to be perfect, so long as it allows users to play all the big Wii titles then that'll suffice.

As for the DD question forcing the BC issue, that's a none starter for Nintendo. With the Wii, all purchase information is stored client side (hard to believe I know, but this is Nintendo we're talking about here), purchases aren't tied to any sort of online account. You'll have to keep your Wii around if you want to play your VC titles, even if its successor is fully capable of running that software.

As for the comment about Nintendo needing a "Live style online service" in order to compete in more traditional markets, absolutely and that goes without saying. In interviews regarding the 3DS they've outright acknowledged this fact, so its not something I'd worry about. There's plenty of viable online matchmaking and friends services these days, PSN, Steam, Live, Impulse, heck even Apple have their own gaming service these days, its not something that's impossible to replicate, quite the contrary. Since Nintendo will be offering it for free it doesn't have to be as good as Live (though we already have free alternatives like Steam that show its possible) just "good enough" like the PSN is.

 

[upnorthsox]

I'll post this here as the last few slides give a Crytek view of the console future.

http://www.crytek.com/sites/default/files/ChinaGDC_Xiaomao_submission_FINAL.ppt#450,40,Heterogeneous system

 

[upnorthsox]

Late 2011 for Nintendo.

Late 2013 for Sony and Microsoft.

Would Nintendo really need to include dedicated hardware in order to make BC work on the sort of hardware we're suggesting? The gap between the Wii and its successor is going to be huge in terms of raw power no matter what we end up with and AMD can always make slight tweaks to the GPU hardware in order to make BC less of a challenge (although I doubt even this will be necessary considering similar PC GPUs are managing to render Wii games in HD with supersampling just fine). Heck, is sticking the Wii's PowerPC core on die really going to affect costs all that much? That thing is going to be tiny at 28nm and not have any real noticeable effect on die size, but again I doubt dedicated hardware like that is even needed to provide an adequate BC solution. Software emulation will more than suffice and its an area that Nintendo already have a big investment in. I'm sure they'll already have an internal x86 based Wii emulator, it'll just be a case of optimising it and ensuring compatibility with most major releases. It doesn't have to be perfect, so long as it allows users to play all the big Wii titles then that'll suffice.

As for the DD question forcing the BC issue, that's a none starter for Nintendo. With the Wii, all purchase information is stored client side (hard to believe I know, but this is Nintendo we're talking about here), purchases aren't tied to any sort of online account. You'll have to keep your Wii around if you want to play your VC titles, even if its successor is fully capable of running that software.

As for the comment about Nintendo needing a "Live style online service" in order to compete in more traditional markets, absolutely and that goes without saying. In interviews regarding the 3DS they've outright acknowledged this fact, so its not something I'd worry about. There's plenty of viable online matchmaking and friends services these days, PSN, Steam, Live, Impulse, heck even Apple have their own gaming service these days, its not something that's impossible to replicate, quite the contrary. Since Nintendo will be offering it for free it doesn't have to be as good as Live (though we already have free alternatives like Steam that show its possible) just "good enough" like the PSN is.

2011 wouldn't surprise me a bit for Nintendo but I'd be shocked if MS and Sony let them have a 2yr run all by themselves. Once you start hearing Nintendo rumbling (spring?) one of the others will start to stir and then the other will follow

 

[brain_stew]

Microsoft have invested too much into Kinnect to launch a new console any time soon. If Kinnect flops and fails to drive hardware sales, it'll be hard enough to justify a successor to sharehoders in 2013 let alone 2011. Microsoft have been talking about postponing the Xbox 3 until a ~2015 timeperiod lately, not bringing its launch forward to 2011.

Sony have lost to much money on the PS3 to ditch it any time soon, they'll milk that thing as much as they can before any successor is launched and anyway, Sony's hardware sales are on the increase and they've got the most "future proofed" hardware.

I just don't see how its feasible for Sony or Microsoft to launch a console earlier than late 2012, they're just not setup or ready for it yet, 2013 definitely seems the most likely. Its why Nintendo are in a perfect spot to launch their console in 2011, they'll have 2 years as the tech leader with the best console version of 95% of third party output even if they're very conservative with their hardware upgrade.

If they can continue to attract the casual crowd that they won over with the Wii (and with their software output, they definitely will), its a nice spot to be in. Software that covers all market segments, the best technology, the best first party output, the strongest brand, the best version of all third part releases and sub $300 hardware that they don't have to take a bath on.

 

[Rangers]

Microsoft have invested too much into Kinnect to launch a new console any time soon. If Kinnect flops and fails to drive hardware sales, it'll be hard enough to justify a successor to sharehoders in 2013 let alone 2011. Microsoft have been talking about postponing the Xbox 3 until a ~2015 timeperiod lately, not bringing its launch forward to 2011.

Sony have lost to much money on the PS3 to ditch it any time soon, they'll milk that thing as much as they can before any successor is launched and anyway, Sony's hardware sales are on the increase and they've got the most "future proofed" hardware.

I just don't see how its feasible for Sony or Microsoft to launch a console earlier than late 2012, they're just not setup or ready for it yet, 2013 definitely seems the most likely. Its why Nintendo are in a perfect spot to launch their console in 2011, they'll have 2 years as the tech leader with the best console version of 95% of third party output even if they're very conservative with their hardware upgrade.

If they can continue to attract the casual crowd that they won over with the Wii (and with their software output, they definitely will), its a nice spot to be in. Software that covers all market segments, the best technology, the best first party output, the strongest brand, the best version of all third part releases and sub $300 hardware that they don't have to take a bath on.

Nintendo would not even be able to put forth a box with more power than PS3/360. PS3 and 360 are just barely profitable now, at 299/(299/199, lets call it 249). Plus those consoles are still large, still hot.

Nintendo has no stomach at all for hardware losses or even break even now imo.

Nintendo is becoming more like Apple, more of their profits come from hardware than software I suspect. Especially in the DS case. So reducing their hardware profitability would strike hard at their core. In order to get a box more powerful (let alone significantly so) than PS3, at 249 or less, with good profitability (and still way less than Wii) not happening.

I forsee a pretty grim future for Nintendo in consoles quite frankly, the Wii itself is in some doldrums (selling 15k/week in Japan, faltering in USA), and I dont see a good path forward for Nintendo (3D will be done better by Sony and MS, and motion will be done better by Kinect, so what gimmick is left next gen for Nintendo?). OTOH I think the 3DS will be a smash.

 

[Mobius1aic]

Considering the huge numbers of casual Wii players, if Nintendo wants them to transfer to the next Wii system, they'll need a Live like system that practically spoon feeds them, and I think the current channel system could be implemented to work well within that aspect. As for hardware profitability, Nintendo could probably make an easy profit per system with a current Athlon II quad core 45 nm AMD chip, Redwood GPU, 1 GB GDDR5 and selling it for under $300. Of course it would be best to wait for Fusion for the sake of flexibility and cost. A very cheap case & PSU combo + cheap AM3 mobo + AMD Athlon II x4 + Radeon 5670 1 GB + DVD drive + HDD + wifi card can be had for ~$300 (notice I left out an OS and system RAM). In a console much of the material costs could be excluded due to all the extra components that would otherwise be on the mobo and graphics card. But it pretty much proves that Nintendo could get off the shelf PC components now, properly amalgamate them by cutting out unnecessary components, customizing the interfaces + design and still sell a system for under $300 and still make a profit.

 

[Rangers]

Nintendo sells the Wii for $199 currently. What PC components would that be equal too? Pentium 2? Geforce 3? You see Nintendo is not interested in any realistic pricing at all.

Nintendo could get off the shelf PC components now, properly amalgamate them by cutting out unnecessary components, customizing the interfaces + design and still sell a system for under $300 and still make a profit.

The CPU is $100 and the GPU is $100. This will be selling wholesale for $280 or less, so for that you need a motherboard, hard drive (there goes your $80 for those two items) profit, power cables, controller, assembly, shipping, motion control stuff?. It's a tight fit to put it kindly.


And what would be the point again? Getting blown away in 2-3 years when MS and Sony are ready? Or do you think this would just have Wii control again? If it did, why would any third parties support it when they dont support the Wii, and when you're talking 3-5 year development cycles, you know MS and Sony are going to have better hardware anyway by the time your game is done?

Basically you need to look at any Wii successor as a total system imo. Are you proposing it be a simple normal console with no motion control (obviously not)? If not, what gimmick are you proposing?

Not to say there isn't one out there I havent thought of. I didn't see the 3DS coming.

 

[Squilliam]

They can certainly get away with (alot) less than cutting edge and still deliver a significant upgrade. There goal should be to get to a rough content object parity so devs don't need separate content bases like this gen, things like advanced lighting, AA, OMB, etc... can all be watered down if need be.

So long as what they deliver is more powerful than the Xbox 360 / PS3 with say 1GB of RAM for instance they'll be able to reap better ports of current generation titles and adequate ports of next generation titles.

If you consider Ontario with it's 80SP and 2 Bobcat cores @ 74mm^2 on 40nm TSMC then they could obviously achieve a much greater level of performance on the same process or on 28nm. If they were to go for something like quad Bobcat plus 320-640SP and a relatively spacey ~15MB on die cache they could probably get that all under 200mm^2 on the TSMC 40nm process or ~130-150mm^2 on the 28nm process whenever that eventuates.

For memory they probably wouldn't need to go for more than a 64 bit bus and low clocked GDDR5. With 4 16 bit 2Gb modules they could achieve 40-50GB/S easily which ought to be more than enough.

As for storage, if they used Blu Ray they could easily double the optical drive throughput of the PS3 by using a 4* Drive and triple it with a 6* Drive. Finally all they would need would be something like 8GB of onboard flash to round out the console.

Theres no reason why they wouldn't take what is effectively an uprated Arcade and charge more for their console. They'll probably be supply constrained initially anyway so charging $299 for their base console could net them a lot more revenue especially as their internal component pricing comes down over the intervening years between their launching their console and Microsoft and Sony doing likewise. Even with powerful components they can still have a console which has effectively much higher margins.

 

[liolio]

For some irrational reasons I don't expect Nintendo to come come with some "standard" hardware for their next console.

I agree with Squilliam pushing out something good enough in 2011 is more than achievable. It makes me think of the thread "Xenos: how fast is it?" Nobody dared but an obvious answer is "by nowadays standards dog slow...".
The same kind of comment raises in my mind when I think about the last xbox revision and its unified chip. Whereas MS/IBM beat AMD to it the overall chip characteristics are nothing to wow at.
Low transistor density, low perfs per mm², too high Watts per mm².

 

[MarkoIt]

I think Bobcat is really interesting. It's a huge leap forward from a perf/mm^2 and perf/watt POV, if AMD claims are true (90% of the performance of a K10 core).
At 40nm, it takes just about 10 mm^2 for the core with the L2 cache.. how many core could fit in a 150+ mm^2 and 50-75 W chip at 20-22nm? 16-24?
It seems to me that 16+ threads is what the game industry is asking for the next gen, and this could be the cheapest way to provide them. A "Many slow threads" philosophy..

 

[czekon]

I dont think kinect is something that hold back MS to realase next gen. Kinect can be compatible with new hardware.

 

[Rolf N]

Late 2011 for Nintendo.

Late 2013 for Sony and Microsoft.

Would Nintendo really need to include dedicated hardware in order to make BC work on the sort of hardware we're suggesting? The gap between the Wii and its successor is going to be huge in terms of raw power no matter what we end up with and AMD can always make slight tweaks to the GPU hardware in order to make BC less of a challenge (although I doubt even this will be necessary considering similar PC GPUs are managing to render Wii games in HD with supersampling just fine). Heck, is sticking the Wii's PowerPC core on die really going to affect costs all that much? That thing is going to be tiny at 28nm and not have any real noticeable effect on die size, but again I doubt dedicated hardware like that is even needed to provide an adequate BC solution.

They could let programmers use both the tiny legacy processor and a more modern architecture side-by-side. Nintendo already did that with the DS (ARM7 for GBA/DS coprocessor, new ARM9 core for DS games). It gives them cycle-perfect compatibility with old software, but without the clock constraints of sticking with the same architecture outright.

I don't know how high a G4esque PPC could be clocked on a modern process, but whatever the answer is, more recent architectures would surely go higher.

 

[Prophecy2k]

I forsee a pretty grim future for Nintendo in consoles quite frankly, the Wii itself is in some doldrums (selling 15k/week in Japan, faltering in USA), and I dont see a good path forward for Nintendo (3D will be done better by Sony and MS, and motion will be done better by Kinect, so what gimmick is left next gen for Nintendo?). OTOH I think the 3DS will be a smash.

Erm... Have you been following all the gaming media on the device?!?! :-S

The way things seem to be shaping up for the device at the moment, I certainly don't think so...

Imho I don't even think Nintendo will gun to put out a performant Wii2 in 2011 if their competitors are gunning for 2012-2013. As much as they potentially could do something powerful "enough" for a Ninty console and the next gimmick they try to push, they have no reason to try to force a new generation until MS & Sony are ready to upgrade their boxes too...

I mean the biggest factor that Ninty would need to get right next time would be their third party support. It practically caved completely for the Wii this gen, and for no other reason than simply the box just wasn't powerful enough for third parties to port their big multiplatform titles from the PS360 in any kind of economically feasible manner (those who did do Wii ports had to assemple entire teams to do the work).

Let say they put out a console in 2011 that's just a bit better than the PS360. So when the PS4 and NeXtbox release in 2013 with radically different archetecture and oodles of computing power over the Wii2, then Nintendo will simply be in the exact same position they are in now. By releasing 2 yrs earlier than the rest with a box barely different in power from the current console powerhouses, they won't be providing the core gamer with enough of an incentive to upgrade to the Wii2. So when the PS4/NeXtbox release and they're the shiny new computing juggernaughts that everyone expected them to be, then all third parties as well as the core gamer userbase will migrate over to those leaving Ninty in the exact same position as they are in now with the Wii.

It's best for Nintendo to wait. They could even throw out their own fancy new peripherel to try and extent the Wii life that little bit more until 2012-2013 when the rest are ready. That way they can all launch again on almost equal footing and they'd not only be able to capture the third party support but they'd also have the strength of their Nintendo brand helping them with the casuals (which i personally believe will play a much bigger part in capturing the casual and non-gamer userbase next-gen )

 

[kagemaru]

kagemaru - Who's talking about tech parity? We're talking about Nintendo making a system more powerful then 360\PS3 in 2011. To be honest I can't see a reason to make a system less powerful then those two when backwards compatibility is no longer an issue.

I thought Squilliam was making an argument to how Nintendo would have benefited more if they released a system that was up to spec this generation:

Because being below the tech parity has cost them in many ways....

Yeah Nintendo can, and probably will release a system that's more powerful than the 360/ps3. I just don't see much point if Sony's and MS' next system will just beat it out again in the performance department.

 

[liolio]

Performances are far from telling the whole story. The Wii would have done better if it had decent performances. Most likely its market shares would be ps2 level. Who cared about the fact that the ps2 was not the most performant system?

I don't think that Nintendo will release a new console in 2011 anyway.
Still a launching before the others would be a smart move. Say a "non world wide launch" (so Japan) during H1 2012, with the system available everywhere for fall 2012.
Nothing says that one extra year or a bit more would allow Ms or Sony to crush Nintendo system perf in a way that perceived to mass market.
Right now I can't see how a person really concerned by perfs can pass on PC. I don't think it will be different in two or three years.