Wii U hardware discussion and investigation *rename

Status
Not open for further replies.
But now everything has settled down to sending triangles to a GPU to be rasterised and shaded using conventional, years old shader languages.
Sorry, but I strongly disagree with the "settled down" part. Yes, perhaps the changes are not that apparent nowadays, but computer science never stops or stalls, and just like before, things still get faster and nicer every day.
 
Unless there is something in there that lets things be done in a fundamentally different way than 360/PS3, then I wouldn't expect huge strides over the launch titles.

Most of the 360/PS3 improvements have been core rendering algorithms and understanding the production process at that scale.

The same learning curve will exist for PS4/720.

Sure people will get a better handle on what works and what doesn't with WiiU, but I wouldn't expect huge strides.

For the part in bold, there are definitely some "Nintendo-patented" things in the GPU. It's just not known what they are exactly.
 
For the part in bold, there are definitely some "Nintendo-patented" things in the GPU. It's just not known what they are exactly.

Sure I'd be surprised if there weren't.
But I wouldn't expect anything game changing, Nintendo just isn't at the forefront of 3d graphics.

Honestly I was stunned how poor Nintendo's E3 showing was, they were launching a new home console for god's sake.
There was not a single WOW moment in it.

I actually think the bigger issue is that we're likely to see a 2013 console demonstrated before they ship.
 
I worked on launch-window title for Wii, it wasn't any different than Gamecube.
Well, of course. Wii however is rather an exception to the norm though, wouldn't you say? I don't think there's ever been a single console ever where the next generation (if that term even applies to the Wii) was as closely resembling in hardware as with the Wii... Even the new I/O options aren't even handled by the original chipset, there's "starlet" for that stuff. It was just kludged on, with minimal effort spent to make it happen.

Was our next title better? Yes. Was it because we discovered new stuff about the machine? Not really. It was because we had time to iterate.
Well, that's pretty much what I'm saying though, isn't it? It's not as if you would be discovering entirely new execution units and texture pipelines and stuff, but rather learning how to get things done more efficiently...through iteration, or by experience if you will.

Launch titles are looking underwhelming according to some.
Months ahead of actual launch though. Things might well change until november-ish.

I'm suggesting that believing the first one is dumb. That leaves the second and third options. You seem to be suggesting the second.
I suggest sorta cross between one and two. Not that devs are incompetent... If they were, they wouldn't be devs (or would be developing OS code for microsoft, heh! ;)). But certainly devs must always be inexperienced with a new piece of software. You put Michael Schumacher in a new F1 race car, it's not as if he's never driven a F1 racer before of course, but this one behaves differently compared to the previous, subtly perhaps, but enough so that even a world-class driver needs time to adjust and adapt for it.

I think that's a little optimistic myself, but if one is a Nintendo fan they could be forgiven for choosing optimism over history.
Lol, well, I am a Ninty fan I guess. Not as big as some for sure, but NES was my first console. Got it even before I had a computer. Actually started my Nintendo carreer with an orange Donkey Kong Game&Watch while I was just in third or fourth grade.
 
* Maybe the developers are incompetent and/or unfamiliar with the hardware..................
..............I'm suggesting that believing the first one is dumb.
With all due respect, that might be not that dumb all the times.
Back in the early days, we were working on a project with a new hardware, and I (as the director) had an idea, and asked the programmers to check if the system could handle such visuals with 60fps or not, so we could decide if it's possible to go into that direction with the development.
Four very good and experienced coder veterans - with dozens of finished project behind them - were working hard for 3 month and tried everything they could, but they were not able to pull it out faster then 15-20 fps, which I could not accept because it's just not fun to play with a slideshow. These were all good engineers, and I honestly respected and liked them a lot, but somehow I knew that I' did not ask something impossible.
So when there was no more time, I kindly asked them to let my good old friend see what they achieved so far, one more opinion can't hurt much, and perhaps the wisdom of crowds could solve our problem. My friend was working on a completely different project (same company, but different building), so he had no idea whats going on, but he was a true genius and we were talking (and drinking =]) together every night anyway, so we agreed that they tell him everything on Friday, and he will spend the weekend with the devkit and see what can be done.
The s**t was running fluid 60fps on Monday morning. The other four guy was all exceptional programmers, and they never had problem to code anything, but sometimes you need something more. It's the same story with every proffession from the carpenters to brain surgeons, and programmers are no exception. It's possible that 100 out of 100 can get the job done flawlessly, but only 1 out 100 or 1000 will be able to find the optimal solution... only one will have the gift to spark the right idea in his head and you can have the other 4 or 20 in the room for years, they won't do it for you.
 
I acknowledge your stance, Shifty, but I still believe my point stands well founded. :) Rendering pipelines and whatnot aside, at the end of the day you still have to shoot for a specific hardware performance target, and aiming too high means a lot more extrra work than aiming too low.
Okay, that's a somewhat different take to my argument. You're saying that the titles on Wuu could be aiming lower than the performance of the machine because devs have been playing it safe. I don't think that stance stands though because if they aimed well below the machines specs, they could enable AA and AF and get higher framerates for 'free'. The absence of any AA in Wuu is a clear indicator to me of a low performance GPU.
 
Sorry, but I strongly disagree with the "settled down" part. Yes, perhaps the changes are not that apparent nowadays, but computer science never stops or stalls, and just like before, things still get faster and nicer every day.
Whatever changes happen, I expect them to be pretty universal across the industry. Maybe there's a small chance of a raytracing GPU at some point, or fully programmable Larrabee type GPCPPCGPPUs, but the moment someone can present a clear advantage, even else will join in (like unified shaders).
 
Sure I'd be surprised if there weren't.
But I wouldn't expect anything game changing, Nintendo just isn't at the forefront of 3d graphics.

Honestly I was stunned how poor Nintendo's E3 showing was, they were launching a new home console for god's sake.
There was not a single WOW moment in it.

I actually think the bigger issue is that we're likely to see a 2013 console demonstrated before they ship.

Oh yeah. I'm not expecting anything game changing either. Just it showing it's noticeably more capable than PS360.

As for E3 (and last) they've done a horrible job communicating Wii U. For example if they had just said they were focusing on launch window games at the beginning of the press conference, I think the it would have made it more tolerable.
 
But why make a judgement call on launch titles, especially when there is a good chance most of them were on v4 dev kits? The final didn't give a big performance boost, but the point is why make that type of conclusion right now?

I remember how rough some (most) 360 and PS3 games looked 6 months out before launch but also how titles for the original Xbox were very much WIP. As final dev kits come rolling around and optimizations made a lot of eye candy may have the switch hit and framerates stablize. Now a lot of improvements in tools may be too late for the launch titles the fact is we should wait before passing judgment.
 
I don't think that stance stands though because if they aimed well below the machines specs, they could enable AA and AF and get higher framerates for 'free'. The absence of any AA in Wuu is a clear indicator to me of a low performance GPU.
It might just be that they haven't enabled AA yet. On the 360, AA is essentially free, except you need to tile higher rezzes, but Wuu alledgedly has 32MB eDRAM according to this thread, so that shouldn't be an issue, at least for 720P. I don't think ATI would design a GPU over half a decade ago that could do "free" MSAA, and then not be able to replicate the same feat in the 2011-12 timeframe...

So I would hesitate to draw any far-reaching conclusions from the E3 software. There's simply too many unknown variables at this stage.
 
I actually think the bigger issue is that we're likely to see a 2013 console demonstrated before they ship.

I think the same, actually more, I think that one of them is capable of finding a good "working" kinetic + "traditional" controller solution that would have better gfx, able to use these latest dev/content friendly editors and be more innovative at the same time.


On the other side, one could say all of that about Vita:???::!:

Personally games will win me or not.

Anyway until a clean reliable leak or a good/big dev game (near launch of the game) I dont think we should make any conclusions.

To much contradictory info, specially details.
 
Sure I'd be surprised if there weren't.
But I wouldn't expect anything game changing, Nintendo just isn't at the forefront of 3d graphics.

Honestly I was stunned how poor Nintendo's E3 showing was, they were launching a new home console for god's sake.
There was not a single WOW moment in it.

I actually think the bigger issue is that we're likely to see a 2013 console demonstrated before they ship.

The following concern may also be one of the reasons why Nintendo decided to hold back a bit. They have been known to being a bit paranoid (somewhat justified in some cases), so they may not want to reveal too many games in order for their competition to not feel too threatened by the Wii U. This may be effective as long as third party developers are also within the info loop, and that there really are alot of games that have not been announced.
 
It's amazing to me that will still have little concrete info on Wii U. The CPU is now an alleged triple core broadway with upped clock frequency and 3 MB edram for cache. Still don't know much about GPU other than it has specific customizations made for Nintendo and has 32 MB of edram. It's like it's impossible to get anything juicy on the machine, something tangible that tells us where the weakness of he console lies.

It could be that when some devs said the machine is weaker they were referring to the CPU. I hope the thing at least has decent FPU capabilities added in. It would be like Nintendo to gimp the machine one way or the other.

I'm not concerned with these screen shots considering we are still at least 5 months from launch if they launch in November. Still a good amount of time to improve and polish the visuals to at least an equal level of the current crop. Not expecting miracles for launch games given my utter dismal expectation that Wii U will only be 50% faster. So the games after that is where U expect to see noticeable visual improvements. I also will give Nintendo games some slack considering they're not that experienced with hardware on Wii U's level. Give them time to play catch up.
 
Nobody has said that.
Espresso claimed it is broadway, IE, PPC G3-equivalent or whatever. G4? Something along those lines anyway. What information he based that claim on I have no idea. Ask the guy, maybe it's something better than random forum mutterings from GAF or such... :p

Espresso said:
It isn't power7. it isn't SPU or cell. it isn't a 4xx. It is the same core as Wii, with 3 of them and larger L2's, clocked a little bit faster.
 
So I would hesitate to draw any far-reaching conclusions from the E3 software. There's simply too many unknown variables at this stage.
I don't tend t draw conclusions ever; I'm just leaning more in favour of the existing evidence given what's been said about it. I don't buy the 'AA's not enabled yet' idea. we've had 2xMSAA for donkey's years. 32 MBs of eDRAM, 720p frames, and no AA?? You may be right, but the evidence is strongly in favour of the lower-spec'd machine, both what we've seen at E3 and Nintendo's known preferences for hardware and the various rumours. I'll be very shocked indeed if we end up with a console significantly more powerful than PS360.
 
I'll be very shocked indeed if we end up with a console significantly more powerful than PS360.
I'd be very shocked as well, very shocked indeed. Especially considering those teeny-teensy fans in the picture of an alledged Wuu devsystem. However, that said, would you consider the (rumored!) ~1.5x PS360 figure that was bandied about at times before this year's E3 as a "significantly" more powerful machine? More powerful, yes, but significantly?

IE, what is your rough estimate of what comprises a significant difference? :)
 
'Significant difference' would be perceivable differences in 3rd party cross-platform games, like ME3 rendering at a higher frame, or with better IQ. Just better textures wouldn't count as an overall '1.5x' improvement in processing power. I also wouldn't count 1.5x as a significant advance anyway. That wouldn't be a generational advance, but a late-gen advance, like XBox versus PS2 which I think was well over 1.5x faster.
 
Status
Not open for further replies.
Back
Top