Wii U Speculation Thread of Brains Beware: Wii U Re-Unveiling At E3 2012

Status
Not open for further replies.
The Cell has a lot of crazy functions that aren't really necessary. It can't really be compared to a traditional CPU. It's entirely possible that Cell will beat the Wii U's CPU in some ways, including raw GFLOPs. (In fact, since Wii U probably only has a tri-core, it'll almost certainly lose in that area.) (Also, I only mentioned GFLOPs because people like numbers.)

The Cell will almost certainly beat all the next-gen CPUs (unless Sony uses another CELL) on single-precision (ie 32 bit) GFLOPS, although if the Wii's CPU is even slightly similar to Power7, it could well smoke the Cell's double-precision (64 bit) GFLOPS.

These numbers are largely theoretical, though, and in the real world I'd expect the Wii U's CPU to be noticeably more powerful in most general cases, and somewhat less powerful in some specific cases which benefit from the Cell's architecture. I'd also expect it to be far easier to optimise code for, which should satisfy developers' main interest.
 
this is Shelk

lrUTl.png

...ShUlk
 
Paid DLC on the Wii U has to mean some sort of account system.

Publishers will want multiple accounts in a household having to buy DLC separately.

An account system is likely, they're even moving towards it on the 3DS.
But I dunno about having to buy it separately for each account on a single system...
 
The Cell will almost certainly beat all the next-gen CPUs (unless Sony uses another CELL) on single-precision (ie 32 bit) GFLOPS, although if the Wii's CPU is even slightly similar to Power7, it could well smoke the Cell's double-precision (64 bit) GFLOPS.

These numbers are largely theoretical, though, and in the real world I'd expect the Wii U's CPU to be noticeably more powerful in most general cases, and somewhat less powerful in some specific cases which benefit from the Cell's architecture. I'd also expect it to be far easier to optimise code for, which should satisfy developers' main interest.
Double precision units are the first thing to go out when designing a gaming CPU. Double precision doesn't have use cases in games. IBM's VMS units are way too enterprisey for gaming.

That said, if IBM doubles the capacity of their VMX units for the Wii U CPU, which isn't impossible given the shrink from the 360's 90nm to the Wii U's 45nm, they'll already top the PS3's benchmarked performance. Could be nice if GFLOPS matters at all.
 
I'm sure third parties would love to make it so each account has to buy it, but I still don't know if that will work with Nintendo.
And that's even assuming you can have multiple accounts on a Wii U system.
 
Double precision units are the first thing to go out when designing a gaming CPU. Double precision doesn't have use cases in games. IBM's VMS units are way too enterprisey for gaming.

That said, if IBM doubles the capacity of their VMX units for the Wii U CPU, which isn't impossible given the shrink from the 360's 90nm to the Wii U's 45nm, they'll already top the PS3's benchmarked performance. Could be nice if GFLOPS matters at all.

I'll admit I'm not a game programmer, but I would imagine there would be certain aspects of physics simulation at least that would benefit from double-precision (for example fluid simulation)?
 
I don't see Nintendo splitting it up like that. It would expect it all be tied to a console, or all tied to an account.

Then it will be via console, no way are Nintendo going to lock MK DLC(for example) because the wrong Mii is player one, they only reason I'm not sure about 3rd parties is because I suspect EA & Ubisoft will try and push Uplay/Origin to be a visible presence on the Wii U.
 
I'll admit I'm not a game programmer, but I would imagine there would be certain aspects of physics simulation at least that would benefit from double-precision (for example fluid simulation)?
I'm not a game programmer either (major in computer science though). Double precision is needed for accurate results in many applications when (a series of) tiny rounding errors can have profound effects, e.g. in medicine or engineering. In computer graphics you just want your scene to look decent enough as quickly as possible. Although not as accurate as DP, SP should still produce great physics. Although there may be errors in calculations, they are on the order of many decimals behind the decimal dot, which is impossible to notice on screen.

So SP physjcs is what gets the job done and is much faster at that as well. That makes DP obsolete for gaming.
 
I wonder what Nintendo's internal focus is on in terms of graphics. Is it current gen games at 60 fps? Super fine IQ? What is this system designed to be strong with?
 
I'm not a game programmer either (major in computer science though). Double precision is needed for accurate results in many applications when (a series of) tiny rounding errors can have profound effects, e.g. in medicine or engineering. In computer graphics you just want your scene to look decent enough as quickly as possible. Although not as accurate as DP, SP should still produce great physics. Although there may be errors in calculations, they are on the order of many decimals behind the decimal dot, which is impossible to notice on screen.

So SP physjcs is what gets the job done and is much faster at that as well. That makes DP obsolete for gaming.

My reason for mentioning fluid simulation in particular it is that some of the techniques I've used for similar simulations can exhibit instability as a result of precision errors. That is, it's possible for the simulation not just to be slightly less accurate under lower precision, but to actually "blow up", for example with the fluid starting to vibrate at a higher and higher frequency with no external force acting upon it. As I say, though, I'm not a game programmer, so I don't know exactly which techniques are used for fluid simulation in games or how differently they may react in single-precision compared to double-precision.
 
My reason for mentioning fluid simulation in particular it is that some of the techniques I've used for similar simulations can exhibit instability as a result of precision errors. That is, it's possible for the simulation not just to be slightly less accurate under lower precision, but to actually "blow up", for example with the fluid starting to vibrate at a higher and higher frequency with no external force acting upon it. As I say, though, I'm not a game programmer, so I don't know exactly which techniques are used for fluid simulation in games or how differently they may react in single-precision compared to double-precision.
I don't know much about fluid simulation, but I guess such complex models of fluid systems are out of the question anyway. Even if they still mess up, then the design choice would probably be too change the model to make it look right. In video games, it doesn't have to be a simulation if you can make it look like one.
 
I wonder what Nintendo's internal focus is on in terms of graphics. Is it current gen games at 60 fps? Super fine IQ? What is this system designed to be strong with?

We don't know untill we figure out what they are developing and which studios are behind it. If Reggie's talk about Wii U being for the 'core' gamer with lots of disposable income is true they'll create different type of games for that audience. So it can be anything between AAA HD Metroid, Zelda and new bankrolled Nintendo published third party titles titles to upscaled Wuhu Islands and Nintendogs.

We do know Retro Studios hired some smart people who are great at rendering large open world environments and Nintendo set up a new technology group in the US headed up by one of the PS3 RSX engineers. I think we'll see lots of surprising stuff coming from them but I guess the difference between graphics quality between titles is going to be even more dramatic than on Wii because studios like Retro Studios can get insane results while Animal Crossing and Mario Kart will keep their current style. We already had a taste of this with the difference between the Wii U Zelda and Garden demo's compared to something like 'Find Mii'.
 
My reason for mentioning fluid simulation in particular it is that some of the techniques I've used for similar simulations can exhibit instability as a result of precision errors. That is, it's possible for the simulation not just to be slightly less accurate under lower precision, but to actually "blow up", for example with the fluid starting to vibrate at a higher and higher frequency with no external force acting upon it. As I say, though, I'm not a game programmer, so I don't know exactly which techniques are used for fluid simulation in games or how differently they may react in single-precision compared to double-precision.
The only point accurate fluid simulation has in game programming is if a game is entirely designed around fluid physics, and even there chances are it wouldn't be based around anything realistic enough to require high precision. It's extremely rare that a gameplay aspect is based on reality, most game development is based on finding the easiest way to make something fun, and reality just doesn't give a good bang for your buck.
 
I wonder what Nintendo's internal focus is on in terms of graphics. Is it current gen games at 60 fps? Super fine IQ? What is this system designed to be strong with?
A good measuring point is to look at the 3DS. The 3DS is capable of high quality graphics using shaders, but Nintendo's first-party games tend to look like Wii games. So I'm betting Nintendo's internal focus for their own games will be high definition Wii-style graphics without any compromises (meaning, high framerates).
 
You wouldn't imagine the shortcuts the physics engines used in games take for simplicity and performance.
 
I would settle for 720p and 30fps for most games if I could get absolutely phenomenal textures and some great shadowing in games...I hate that pixelated shit so many games try to get away with...wtf is up with the shadows in Skyrim when you are near a candle...seriously.
 
It's bgassassin's funny and roundabout way of saying wiiU will have esata. For sure.

:) thanks.

USB 2.0 theoretical bandwidth is very small compared to USB 3.0 (and eSATA). Since Nintendo is allowing the usage of external HDDs, the idea is to avoid bottle-necking the HDD as little as possible.

So we could install the game to the EHDD and stream/load from it? If so then the so-called casualz can play from the disc and the so-called hardcorz can install onto their EHDD and play from that [with the disc in the console presumably] and everyone is happy. That is assuming the rumour about a small amount of internal flash and optional USB EHDD support is correct.

Do we know what the costs would be for Nintendo to include USB 3.0 and/or eSATA as distinct from USB 2.0? I imagine that the former would be more expensive for them than the latter since it is newer?
 
I would settle for 720p and 30fps for most games if I could get absolutely phenomenal textures and some great shadowing in games...I hate that pixelated shit so many games try to get away with...wtf is up with the shadows in Skyrim when you are near a candle...seriously.

Agreed. I don't recall where (maybe the AV Forums website?), but there is an outstanding chart depicting how close to the TV set one would have to be in order to tell the difference between 720p & 1080p - and the distance is incredibly small. I'd much rather see that processing power put to other use..
 
I think we'll see Nintendo farming out some of their titles/franchises to western third party developers to cater to the US market mostly. Iwata already kinda hinted at this in the recent investors Q&A's and he also said Nintendo was willing to invest. They need to get some serious core development going if they're really going to compete with Sony and Microsoft head-on.
 
I disagree, I can tell the difference between 720p and 1080p from 5 feet away from my 40" TV, it's very obvious. And as a next-gen system, I think it needs to do 1080p very well.
 
I would settle for 720p and 30fps for most games if I could get absolutely phenomenal textures and some great shadowing in games...I hate that pixelated shit so many games try to get away with...wtf is up with the shadows in Skyrim when you are near a candle...seriously.

30 fps is ass. I'd take a hit on graphics if it meant 60 fps.

But I do agree with the difference between 720 and 1080p being miniscule. Again, if I'd gladly take 60 fps at 720p over 30fps at 1080p.
 
30 fps is ass. I'd take a hit on graphics if it meant 60 fps.

To me, it depends on the game. I expect nothing less for every Nintendo game. They have a simple enough art style to do 1080p and 60fps. But I would settle, emphasis on settle, if it meant some awesome textures and shadows. Those two things are what throw me off the most when I'm trying to suspend disbelief in a realistic game world.
 
I don't know much about fluid simulation, but I guess such complex models of fluid systems are out of the question anyway. Even if they still mess up, then the design choice would probably be too change the model to make it look right. In video games, it doesn't have to be a simulation if you can make it look like one.

Well, anything that looks like a simulation is still a simulation, just a less accurate one.

There are two games this gen I can think of that do decent fluid simulation, Hydrophobia and From Dust, and I don't know what techniques either use (very different ones, from my impression), but neither seem like the sort of games where the CPU is doing a whole lot else. It's something that I'd expect to see decent improvements in next gen, like we saw in rigid-body and ragdoll physics from the last to this generation.
 
I agree. With the wiimote the appeal it would have on gamers and casual audiences was obvious. Not so much with the tablet. Dual screen technology is nothing new, casual audiences have seen it before, and it hasn't proven to be anything super revolutionary. At least not in the way the wiimote was with motion controls. It works with some ideas, and is a total distraction with others. Heck, the fact that Sony - who is known to blatantly ripoff Nintendo's good ideas - is sticking with one large high-res touch screen instead of going with the dual screen method is proof that the DS line's dual screen isn't a game changer. I don't think the Wii U's dual screen concept will be much of one, either.

That's why I'm hoping that Nintendo is treating the controller as secondary to making a truly compelling console. Hardware first, controller second.
I think it's kind of silly to say that, given the popularity of the iPad and... well, not so much other tablets, but tablets in general.

WiiU has always been, in my mind at least, a partial response to Apple and the iPad. I do agree that, based on what we know so far and based on Nintendo's past history, the WiiU probably won't be terribly well received. It needs to compete reasonably well against the iPad, because there is simply no way that it isn't going to be compared to that. Bare minimum, that means a multitouch screen with a better resolution.
If I were Nintendo, my strategy would be to basically trying to beat Apple to the punch when they inevitably bring iOS to the living room in a big way (as opposed to the "hobbyist" AppleTV right now). if those rumors about Apple bringing out a TV are true, then the games are inevitably going to follow, and not just shit like Angry Birds. Nintendo has more to fear from Apple at this point than MS or Sony.
 
if those rumors about Apple bringing out a TV are true, then the games are inevitably going to follow, and not just shit like Angry Birds. Nintendo has more to fear from Apple at this point than MS or Sony.

But Sony already has TVs.
How does Apple coming out with TVs hurt Nintendo?
 
But Sony already has TVs.
How does Apple coming out with TVs hurt Nintendo?

It's more about really bringing iOS to the living room more than anything else. Whether they do that with TVs, set top devices or (probably) both matters very little. Right now they have a half ass version of iOS on AppleTV, but playing games and such on it is a pain.
 
[Nintex];33764684 said:
I think we'll see Nintendo farming out some of their titles/franchises to western third party developers to cater to the US market mostly. Iwata already kinda hinted at this in the recent investors Q&A's and he also said Nintendo was willing to invest. They need to get some serious core development going if they're really going to compete with Sony and Microsoft head-on.


In terms of 3rd-parties, they already have Monster and Next Level, with EXCITE and Pilotwings by Monster and Strikers, Punch-Out, and currently Luigi's Mansion 2 by Next Level.
 
Yeah, just like the 3DS has to fear the giant boom of mobile gaming.
Uh huh...
Until I see actual proof in sales that this is happening, then it isn't.
 
I think it's kind of silly to say that, given the popularity of the iPad and... well, not so much other tablets, but tablets in general.

WiiU has always been, in my mind at least, a partial response to Apple and the iPad. I do agree that, based on what we know so far and based on Nintendo's past history, the WiiU probably won't be terribly well received. It needs to compete reasonably well against the iPad, because there is simply no way that it isn't going to be compared to that. Bare minimum, that means a multitouch screen with a better resolution.
If I were Nintendo, my strategy would be to basically trying to beat Apple to the punch when they inevitably bring iOS to the living room in a big way (as opposed to the "hobbyist" AppleTV right now). if those rumors about Apple bringing out a TV are true, then the games are inevitably going to follow, and not just shit like Angry Birds. Nintendo has more to fear from Apple at this point than MS or Sony.

Oh, I wasn't speaking about tablets in general. I was referring to the tablet controller of the Wii U specifically. Sorry for the confusion.

But I don't agree with the rest of your point. Nintendo should never, ever, ever-ever-ever-ever-ever try to compete with the iPad. That is not their field, and they wouldn't have a shot in all 7 hells. Upad =/= iPad (or any other dedicated tablet device, for that matter). It is a shell with a touch screen. Nothing more.

Nintendo should just stick to utilizing the unique advantages of a touch screen for gaming on a console, and leave it like that. If they enter this gen with some tangent idea of competing with tablets, they will flounder.
 
I think it's kind of silly to say that, given the popularity of the iPad and... well, not so much other tablets, but tablets in general.

WiiU has always been, in my mind at least, a partial response to Apple and the iPad. I do agree that, based on what we know so far and based on Nintendo's past history, the WiiU probably won't be terribly well received. It needs to compete reasonably well against the iPad, because there is simply no way that it isn't going to be compared to that. Bare minimum, that means a multitouch screen with a better resolution.
If I were Nintendo, my strategy would be to basically trying to beat Apple to the punch when they inevitably bring iOS to the living room in a big way (as opposed to the "hobbyist" AppleTV right now). if those rumors about Apple bringing out a TV are true, then the games are inevitably going to follow, and not just shit like Angry Birds. Nintendo has more to fear from Apple at this point than MS or Sony.

Your mind sucks.

Also, better resolution = bye bye, multiple controller support. Hell, it means "bye bye, using the screen for anything beyond 2D stuff and streaming."
 
I would agree that it is a partial response. I don't see how you can rationalize that it wasn't a decision that was affected by the new Tablet market.

If Nintendo's dumb enough to think that they can compete directly with Apple, they deserve to fail.
And even if they were successful in doing so, Apple would just sue them and get Wii U banned in a shitton of countries.
 
They aren't trying to compete with Apple in any area other than time.
The Wii U controller is just that. A controller. Just an input device. Completely useless without a console.
It is not a tablet. It is not a computer.
 
It was certainly influenced by the tablet explosion, but likely also the success of the DS and the proliferation of smartphones throughout the world.

The idea of the "personal screen" is really the fundamental thing here.
 
I disagree, I can tell the difference between 720p and 1080p from 5 feet away from my 40" TV, it's very obvious. And as a next-gen system, I think it needs to do 1080p very well.

I'd like to see it do 1080p in general, but for one reason: while I have yet to see a game where 1080p actually added anything to the gameplay itself, it really makes split screen usable again for a lot of people. Perfect Dark XBLA in native 1080p where each quarter of the screen has almost the resolution of a single player in a Call of Duty game is awesome. And it runs in 60fps while doing it.

An image I whipped up long ago to demonstrate this to someone:

minRk.jpg


I like 1080p more for the implications of restoring the social days of gaming than the fidelity of it.
 
If Nintendo's dumb enough to think that they can compete directly with Apple, they deserve to fail.
And even if they were successful in doing so, Apple would just sue them and get Wii U banned in a shitton of countries.
It's not about competing with Apple really, it's more about getting a tablet like controller out there when that type of device is hot and relevant.
 
I don't see them loading tons of textures into memory...that should help load times.

Nintendo has always engineered their drives to be predictable and fast. On the GameCube, the devkit also enforced the developers to deal with the actual console's loading speeds by having the devkit's HDD throttle the data speed.

They're probably one of the better developers out there when it comes to their expertise on streaming and caching.
 
Status
Not open for further replies.
Top Bottom