I would like to point out a subjective observation: recent generation of 3DS games (3rd if I'm not mistaken), Mario Golf, Luigi's Mansion, Mario and Luigi and Mario vs Donkey Kong in particular all look incredible, with lots of beautiful geometry and effects, much better than even Super Mario 3D Land or Mario Kart 7. Think about it for a minute, then imagine what kind of lookers we'll be getting for Wii U come 2014!
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
WiiU "Latte" GPU Die Photo - GPU Feature Set And Power Analysis
- Thread starter Thraktor
- Start date
- Status
If it's used in the water in any of the Halo games, it would be 3 and/or Reach. The water in those games has several things which could indicate involvement from the tesselator: animated geometry having interesting draw distance fall-off, and dynamic splashes looking unstable if you pause them in theatre and move the camera around them (the geometry offset shading that gets done on the output of tesselators is prone to shimmering for exactly the same reason that textures can shimmer if the LOD bias is too far negative).
Halo 4's water is mostly a static, relatively low-poly surface. It certainly *could* be generated with tesselation due to the fact that it's not always a 100% flat surface (i.e. downhill rivers), but there's not really much reason for them to have done so.
//============
The instabilities with normal maps being applied to geometry in Halo's 3 and Reach is actually probably one of the main reasons we didn't see the 360's tesselator get used much. Those games could afford some shimmering due to the fact that we're talking about water, but the same issues can happen on static objects using tesselation for geometry LOD, and that can look totally disastrous. Fortunately, tesselators and the understanding of their uses have come a long way since 2005, and careful, informed use can now circumvent these issues. Next gen will *probably* see tesselation used much more often, even on the Wii U.
I thought it was Halo 3's water that used tesselation?
Edit: Quick google says it's actually Halo Reach
Edit#2: Actually there doesn't seem to be a consensus regarding whether any Halo uses tessellation from what I can see. I also noticed the post above me now. Took me a while
Edit: Quick google says it's actually Halo Reach
Edit#2: Actually there doesn't seem to be a consensus regarding whether any Halo uses tessellation from what I can see. I also noticed the post above me now. Took me a while
Question wasn't about feature set it was about use in games.
Lol with the trolling comment. This is just silly.
Good idea or not, devs are not supporting the Wii U. They will support the Durango/Orbis. We will see if the market has contracted permanently or not very soon.
Zoramon089
Banned
The point is you tried saying they were the same and passing it off as a fact
They will be used in games the same. That was the question. Twist it however you like.
I thought it was Halo 3's water that used tesselation?
Edit: Quick google says it's actually Halo Reach
Edit#2: Actually there doesn't seem to be a consensus regarding whether any Halo uses tessellation from what I can see. I also noticed the post above me now. Took me a while
Halo Reach does, at least.
http://www.eurogamer.net/articles/digitalfoundry-halo-reach-tech-interview?page=2
Halo Wars used it for terrain as well:
http://www.gdcvault.com/play/1277/HALO-WARS-The-Terrain-of
lostinblue
Banned
Thank you for elaborating. I don't play Halo so... Was kinda weary of what Halo they did say had used the tessellator so guessed for the last.
Viva Pinata is also a known test case for Tessellation on X360.
Problem with X360 tessellation is not so much shimmering issues as it is the fact that it's a resource hog, vertex buffering what has to be done via multiple passes takes RAM, the fact that it needs vertex shader input makes it so that unified shaders are called into the mess and on top of it all, it doesn't support compression, meaning instead of reducing the footprint it would bloat it. To make things worse it's also slow compared to newer implementations. Basically two advantages for modern tessellation are compression and bandwidth; X360 isn't compliant with neither; eats up bandwidth by needing buffers and doesn't make geometry any lighter (quite the contrary).
That means it's only usable in limited and controlled situations.
Precisely.
No they won't or rather weren't.
Capabilities are very different, Wii U is supposed to feature a modern Tesselation unit, so it can use it dynamically in a open world scenario. On the X360 you didn't have that luck, it was a hog that took resources away so you had to be careful, best suited for either games that took place on closed/enclosed areas or applied to limited geometry within the game. 3DS implementation is meant to spare bandwidth/ram space so I'm guessing it's a lot like truform (ATi 8500 implementation), it's a one pass subdivision thing, no dynamic LOD.
And there you have it, 3 very different forms of hardware tessellation.
snowdog1971
Banned
I've said it before but you'll find publishers changing their tune before the year is out. Once the delayed software such as Pikmin 3, Wii Fit, Rayman Legends, Lego City Undercover, The Wonderful 101 and Game & Wario are released, along with Wind Waker HD, 3D Mario and Mario Kart 8 before Christmas with the possibility of a price cut you'll see the Wii U having an installed userbase between 10 and 15m before the end of the calendar year.
The PS4 is going to be expensive, there's no way it's going to be less than $400. The 720 should retail for $399.99 (imo) which is going to do it a lot of favours but we've got to remember that both consoles are going to be supply constrained. They're not going to sell as well as the Wii U, and even if they do each platform holder will struggle to keep enough supply to meet demand.
Publishers aren't going to ignore the Wii U's superior marketshare now that Nintendo have released a console with a standard rendering pipeline that has a similar architecture to the other two consoles. They're not going to leave that sort of money on the table.
Broken Logic
Neo Member
I think you were actually in agreement, but too much misunderstanding meant you didn't realize it. If I'm not wrong, Schnozberry was trying to say "The 360 had tessellators" with the ? at the end implying a silent "Did it not?". USC-Fan then went on to say, yes it did in fact have tesselators, thinking that Schnoz was disagreeing, when I don't think he was.
I'm a little unsure of what you were saying no to when you said "No, its like the x360 pretty much." Could you put it into a fuller context?
blu
Wants the largest console games publisher to avoid Nintendo's platforms.
I'm just wondering, when you make statements like that, is it because:
a) you're well familiar with the Xenos tessellation unit
b) you're well familiar with a R700 tessellation unit
c) you read a technical paper doing a direct comparison between the two
d) all of the above
e) none of the above
?
Whilst I think games are the main issue, I tend to agree with you. Gamers are bored of the gen, so exciting hardware can help kick off a generation. Public and publisher excitement can translate to sales and build momentum. In fact new hardware can to some degree alleviate the need to be truly revolutionary in your game design, at least at first. It's a nice boost off the line.
Yep. I wonder what they're going to do with a year-late Deus Ex port.
Nah, not really. Even when the enthusiasm for the platform wasn't as low as it is now, publishers weren't really on board.
Can someone create a "Van Owen poor trolling attempt generator" or is it not worth the effort?
AzaK said:
I think everyone should start expecting an industry-wide contraction. While it will probably hit Nintendo harder (considering their 7th gen console sold 100 million) I don't think any console manufacturer is safe.
Aside from the fact that this post reminds me a lot of the "wait for *insert PS3* posts from the beginning of the 7th generation, you have to realize that the Wii also sold a heck of a lot and was mostly ignored. Yes, unlike the Wii, this box has the capability to be ported to far more easily, but publishers were woefully incompetent in the previous generation with a Nintendo console that was on fire for 4-5 straight years. What makes you think they'll greenlight projects for what you say is a 10-15 million lead (which is doubtful in itself). Actually, maybe this discussion is better kept for the Sessler troll thread than this one. This is supposed to be about the Wii U GPU. lol
Please see blu's post.
snowdog1971
Banned
I don't think at least a 10m installed userbase is doubtful at all by the end of the calendar year when you consider the potential system selling software I listed. And by the time November comes along the console should be profitable by then so a price cut should be an option.
lostinblue
Banned
This industry always bets against Nintendo, it's bias at it's best.
I reckon in the GC days, Nintendo was turning a profit (bigger than Sony game division, and Microsoft was losing 6 bilion dollars for the whole Xbox 1 fiasco), yet you had this Xbox game magazine infamously suggesting Nintendo should go third party; they were the ones bleeding money though. Even Bill Gates came out saying he was interested in buying Nintendo as late as 2004.
Even now, Sony makes loss after loss, Vita is a wreck and they seem to be going high spec again, making their machine more expensive than it's competitors or sold at a loss puts them at a disadvantage, but it's Nintendo that is "on the verge of going third party" for journalists. (and for the record I don't wish for Sony to bite the dust, but they're mismanaged and have been for years, and going high spec never actually helped them)
It's kinda mindblowing; specially seeing that Nintendo actually has a chance for market dominance or at least performing well; perhaps not with third party's as they're stubborn as shit and do not believe the platform, therefore won't invest and won't publicize their work for it (and it is also true that multiplatform owners probably won't opt for a Wii U version of a multiplatform game, and that's Nintendo's fault because it really can't compete head-on in specs) but fact is most clients aren't "multiplatform owners", otherwise PS2 and other leading platform examples wouldn't have been all that healthy in multiplatform sales, hell PS3 would be a wreck in that regard and it isn't (although it was, at a point). And Nintendo software division and IP's are the strongest amongst the three, hell, they managed to keep the Wii market leader this gen single handedly no thanks to third parties or journalist feedback/publicity.
Their strategy though, is clear, release first and come E3 2013 the competition will be showing their systems like they were showing their own last year, and Nintendo will show games, and it's games that sell hardware, not tech demos, million particle claims, spec sheets and tech the audience feels like reciting a Harry Potter spell whilst repeating; that can't be said enough. They also expect come December, the other platforms to be in a launch window compass where system sellers have been delayed to ensure quality (making it a little bit like Wii U launch window) while most third party products are still multiplatform with current gen or delayed due to small install bases (no developer wants to take the risk of launching on a small install base if they can wait); Nintendo of course is hoping to have system sellers by then, and a smaller pricetag too (which they already have for granted) which makes adoption easier.
First year of sales are always lackluster if a brand launches first. One has to wonder what competitors will do, system sellers are not ready, older consoles are cheaper, have cheaper games and have better libraries; when PS3 launched PS2 kept selling more units for quite a few months and why wouldn't it? when DS launched GBA kept selling more units on a month by month basis too. When Dreamcast launched it barely made a hit on weekly sales for both Playstation and Nintendo 64, because everybody was expecting the PS2, first year sales of X360 were also crap (roughly 6 million), the same as PS3. Nintendo actually still has 16 million forecast (down from 24 million) and already did sell 3 million before January 2013, that's half X360 first year sales in two months. Go ahead journalists and spell doom and gloom, also know that you're as crappy as it comes at your trade and you just might eat crow again, like you did multiple times already.
It's not so bad. Of course people are waiting, these days gamers know that a price drop will come if they wait long enough and often do that with software, game comes out, no one buys it, wait two weeks, do a pricedrop, sales are healthy: people thought they could wait and so they were waiting. That's why CoD and Nintendo published sales are pretty strong right out the door, because informed consumer has the notion he might as well die of old age before they drop prices, so he might as well fork up the cash.w
I don't have a Wii U yet, I know I can wait; therefore I'll do just that (like I did with the 3DS). A lot of people are doing that, helped by the flailing economy that makes it harder to part with hard earned money, and the fact that the industry is now under attack by post-pc devices whose product cycle involves less waiting for pricedrops and instead bases itself on year-on-year refreshes.
It's a better business model too.
oo, I love cars!
Don't forget that when we talk about horsepower we also need to talk about vehicle weight. 140hp is pathetic in a car, maybe you'd get 7 sec 0-60mph, but with 140hp in a motorcycle, due to the low weight, you'd get 3 seconds. Low weight makes your horsepower much more effective. However, on the opposite side of the scale, high HP will allow you to reach higher top speeds (limited by air friction).
So could we say top speeds in vehicles is like bandwidth, and acceleration is like latency? lol
Schnozberry
Member
I used the wrong punctuation as a typo. Both machines have tessellation units. That doesn't make them like one another. I meant to ask him how he could possibly know how similar or different they are.
lostinblue
Banned
He should know they are very different seeing he goes to B3D where developers already elaborated on such unit and it's limitations to everyday use (also against DirectX 11/modern implementation) and he's the type that spends a lot of time reading because he doesn't really grasp stuff (for the record I also read a lot as probably does everyone arguing on these threads and I'm not ashamed, but I believe I have a way better grasp of things hell, at least I know what's apples to oranges), anyway, that or he's just trolling at this point; I'd guess both.
Even if they were comparable, which they really aren't; tessellation adoption depends on the tech, parity and need for it; which wasn't there last gen. It is here for next gen though, and it's use is advantageous.
Time will still tell how it is used, as a lot of features are supposed to be revolutionary and then never get used for one reason or the other (heh, NURBS, CELL, shit in caps lock I'm noticing), but that's why X360 never used it on a regular basis, units this time around are a lot more effective and part of a standard that being pushed for actively by OpenGL, Microsoft, Epic, Crytek, AMD and Nvidia. I already elaborated on how it should be used for easing out multiplatform development and scaling and I think it's sound.
fabricated backlash
Member
Well that's pretty much at a dead end as far as I understand it. Too many customizations to really use any other chip as a good comparison point. Which makes a decisive power evaluation impossible.
So it's back to DBZ and Market Strategy discussions with a glint of console warfare thrown in for good measure.
lostinblue
Banned
I reckon it's still Krillin. What? Krillin (died a lot, but) was the strongest human, and it didn't have the food needs of super sayans:
Sitting at 33W TDP against certainly +150W 1080p 8GB RAM machine behemoths, with other "food" needs.
It's form factor is also smaller.
EDIT: Oyua is Mr. Satan.
Nevar.
tkscz
Member
He's saying the WiiU's tesselator is the same as the 360's. Of which no evidence is provided. So where he got that info is anyone's guess.
I would suggest moving a majority of the posts to the Wii U technical thread minus the fluff about other consoles.
Shin Johnpv
Member
It's a silly troll comment. Because even if they didn't upgrade the tesselator from the RV7xx chip the Wii U GPU is supposed to be based on, it's still more advanced than the one in the 360.
Apophis2036
Banned
Of course they could have never of known in advance but at this point I bet Nintendo are wishing they spent the tablet costs on an extra core in the CPU, another 4GB's of Ram and a more powerful GPU, even a 600 GFLOP GPU would have only made WiiU around 50% weaker than Durango.
They could have called it Wii 2 or Super Wii and included the Pro controller as standard.
Every PS360 port would have been able to run at 1080p or 720p/60fps and many more developers/publishers would have flocked to the system, as is they see it as more of the same PS360 level hardware and are not excited in the slightest about working on it.
I think some people saying that multi platform games will come when the console reaches 15 million are ignoring the fact that some software just doesn't sell on Nintendo systems, we are already hearing about games like Thief 4, Dark Souls 2, Just Cause 3 and The Witcher 3 being exclusive to PS4/720 and these are only early games, it doesn't bode well for multi platform games in the future that push those consoles further and are released in years two and three.
All Nintendo can hope for now is that their big first party exclusives don't disappoint (which is going to be difficult as some people have been waiting almost three E3's to see them) and that they can continue to get the big selling yearly multi platform releases like Just Dance, Fifa, Madden, CoD and AC for WiiU.
If you own or plan on buying WiiU for anything other than Nintendo games and the big selling yearly franchises I think it's going to be a massive letdown.
BlockheadBrown
Member
Waiting on big purchases? Yep. Cashing in on inexpensive games? Yep. I have a feeling the same will hold true on the next consoles from Sony and MS. Maybe not to as large of an extent as what has seemingly happened with the Wii U, but...
No problem. I was answering his question would it be used unlike the 360 were it was only used in select titles. I said no its will be like the xbox 360 and only select titles will use it. Never said anyhing other than that.
Wiiu is clearly more advance but that doesnt change that only select titles will use it features. That was the question.
First of all, that's not what you said. Secondly, you should either ask money for being clairvoyant or stop trolling.
That is because people who game on Nintendo consoles generally prefer things to be "creative" and "dynamic" as opposed to just buffed up bulk and number on top of more of the same old thing.
The I must digrees. Why is the last two pages filled with nothing but Sony and Microsoft comparison. I don't have time to check, but I fell that someone made a comparison or suggested some kind of parity, and others jumped in to defend what they saw as an attack on their preferred companies hardware.
Exactly how much progress have we made on Latte? Its feels like its getting nowhere. There is more talk about the PS4 than the Wii U's GPU right now.
On that note, I have a few questions regarding the GPU capabilities.
1. Is their any form of estimate on the max polygon capabilities? If I remember correctly, the Gamecube could draw polygons with its CPU. Since the core architecture is still their in the Wii U, can the Polygon count be enhanced using both Latte and Expresso in conjunction?
2. Exactly how much benefit would the Shader Modal 4/5 provide over the Shader Modal 3 based GPU's of last gen?
3. Exactly how much of a hit does the Wii U take when running gameplay on both screens? How much difference would it make in performance if a game were made that only used the T.V. screen didn't use the Gamepad at all?
Thanks for telling me what I said. Lol
And people wonder why people dont post in this thread. Fanboy circle the wagons way too much in here. Instead of debating we just call people trolls. Too funny...
Seeing how little info is given to dev i dont think we will ever know.You and I go way back - how bout telling us what the tesselator is like, just for me?
The issue is that you said you do know.
USC-fan said:
So if devs don't know, how do you know how the tesselator will be used?
Typical USC-fan post. Pulling crap out of your ass and then screaming persecution and yelling fanboy whenever someone calls you out on your shit. Fucking delusional.
If you want to make bogus claims, give reasonable support for your arguments instead of crying like a goddamn child.
lostinblue
Banned
Nintendo wouldn't do that even knowing what they know today; in fact I'm everything regarding competition specs is falling onto their projections (taking 8 GB of GDDR5 aside, that is). Their strategy is and always was about differentiating.
With a mere 600 Gflops? No dice, more capable of attempting 1080p, yes; able to do X360 games at 1080p as if it was a switch? not so.
You'd need a 800 GFlop part for that to be a more linear transition.
And current gen port problems when it comes to coding for a different CPU part wouldn't be erased by the addition of an extra core (first and foremost even because current gen code is at best optimized for 3 cores, so they wouldn't tackle that "extra" core). I think yes, they should have gone for it (wouldn't hurt, the thing is very power efficient and each core takes little silicon anyway), extra cores though would have made it more future proof, rather than better for current gen ports.
And as you know, the GPU is more powerful and yet a ATi/AMD part, therefore similar enough so GPU code for X360 should get freebies all the time, yet some developers have been struggling with it, it probably has to do with "beta" dev tools, "beta" documentation and little investment (in people numbers, quality of the personnel, time and budget).
That's a bias that get's repeated so much that it becomes true.
Third parties treat Nintendo systems as second rate priority (and get second rate results) of course consumers should treat them as second rate products.
Happened on the Wii, it was a myth at first when developers were calling it a fad, software was selling, crappy software they did even; then they used that money elsewhere, did late ports or low budget spin-off adaptations with zero publicity, insisted in on-rails bullshit when the community was already pissed off or insisted in "experiences" surrounding doom and gloom "if this on-rails outsourced endeavour and or indie niche product doesn't sell we'll pull the plug!" of course the market collapsed.
Then it became true, I mean: dah.
The market was there at first, thing is they didn't want any Fifa or Madden All-Play bullshit.
People and Devs also fail to understand the whole blue and red ocean thing Nintendo goes on and on about; it basically means that if you do an FPS for the X360 it'll go against Call of Duty, Crysis and Halo, you're as good as done for. Now, lot of FPS's learnt it the hard way this gen, but blame it on timing or end quality of the product, and perhaps that has merit to it; but above all they were also competing in a very saturated market to begin with and fighting swords with sticks at that.
Nintendo is remarkably lacking in some genres, such as serious racing, yet no one ever tries to deliver (and deliver with quality and a marketing budget) on that blue ocean. Instead they try to do yet another platformer, because it apparently makes sense to go up against Super Mario.
That's like saying Sonic All Star Racing probably sells statistically better on X360/PS3, just saying. (and if it's selling on the Wii U that's pre-emptive strike seeing there's no Mario Kart yet and little software available)
Of course it also makes sense to put Okami and Dark Siders on the Wii/Wii U, but it is also competing against more notorious titles for the same thing (and therefore if fans of such genre can get only one title, they specifically won't get those), so it's a matter of whether gamers can and want to absorb more of the same kind of content or not.
Conjecture and balance, basically. But going for the same genre first/second party offerings are good at it's theoretically not as secure as their logic dictates, in fact it can be very dumb.
At one point, developers have to let go of their pride and embrace real multiplatform.
Sure Nintendo could have gone and made things easier, but there's way less leeway for technical impossibilities than last gen. The question really is on whether they want to have their software on it or not, and time and again they're putting their dime somewhere else and not deeming multiplatform even for current generation games as a worthwhile bet for the Wii U, not because it can't be done but because they can't be arsed to do so.
And as always they're bound to pay dearly for it; yet they don't realize it's the principle that counts, it's like realizing that the whole approach to new IP in this industry is very, very dumb. So they do something new and unknown for a saturated market, doesn't sell, marginalize and throw it in the bin.
Now, if the IP/Game is good, you might as well compromise and look at it from the start as multiple games. Chances are second game in the series will sell better as will the third. IP's grow, just like that.
Acceptance for an unestablished IP on Nintendo platforms has the same potential, Tekken on the Wii U looks off, but if they start coming people will have brand awareness for it just the same. Call of Duty on the Wii probably speaks for itself.
Because you've totally seen those pages. (along with the documentation for X360 tessellator unit)
Delusional sounds right.
lostinblue
Banned
You don't debate shit, you throw one liners as if you know anything; except you don't, hence why it's controversial and why you're a victim (of yourself).
The word count of people answering you or "debating" whatever here, and your posts should be illustrative; you want us to write a thesis about your wrong one liner spoke as fact, thesis that that you'll answer later on as "basically what you're saying" (another favorite kind of response of yours).
You must be trolling alright.
JohnnySasaki86
Member
Really? Cause if I remember correctly the tesselator unit in the RV7xx was disabled and unusable. Wasn't this because it wasn't very advanced and couldn't work with modern API's and what not?
One thing thats for sure, is we have no info on whether the unit was modified from the original RV7xx. So the only thing we can say at this point, is it is the same tesselator unit. We don't have any other info yet to say otherwise.
One thing we could logically assume is the tesselator unit in the next gen systems will indeed be more advanced then the one found in Wii U. I don't think its the same as the one found in x360 though.
lostinblue
Banned
I'm probably to blame, but I was mostly having fun.
I'm bad at keeping ontopic, as everyone must have realized by now.
No advances, hence the talk going elsewhere for a while.
They could, every CPU can do it... commonly called a software renderer; but it makes no sense to do so; GC didn't do it in fact, just some manipulation, transformation and tracking calculations.
GC didn't have vertex shaders on the Flipper, just a fixed function T&L pipeline, so those manipulations had to be done via CPU. With a unified modern shader pipeline though, it really makes no sense to use CPU to manipulate vertex.
As for max polygon capabilities, there's a reason why we stopped hearing about those. It stopped mattering all that much with normal mapping and displace maps being all the rage (and now tessellation). It's certainly adequate enough for it to not be a problem.
Quite a bit, it's the difference between being able to do CUDA/OpenCL/CPGPU or not.
Shader Model 3 to 4 difference means going from 512 to 65536 instruction slots which amount to 128 times as many instruction slots, as for vertex it means going from 512 instruction slots to 4096 (a 4 time increase) and for both it means going from 32 temp registers to 4096, again a 128 time increase. Comparison tables here.
Bare in mind though, X360's Xenos was Shader Model 3+.
Depends on what you're doing; the equivalent of split screen, same image albeit mirrored or simpler input gameplay/menu's.
Split screen takes the same hit as doing split screen by regular means, except main screen retains it's resolution and alternate gameplay/different camera get's sent to the gamepad, that and extra eDRAM used up for it, Mirroring is said to be free; simple input/menu's take extra eDRAM but providing it doesn't have 3D it means no Z-Buffer is needed.
If it's a different scene/complex objects/assets altogether, that's bound to hit the most, but it depends on a case by case basis.
lostinblue
Banned
It wasn't disabled, nor "not very advanced" quite the opposite (for the time); ATi tried for the longest time to get Tesselation on DirectX specifications (ever since R2xx/ATi 8500 in fact), but that only happened with R8xx/DirectX 11, so the thing was there but didn't see much use because it was a open "beta" of sorts.
The requirements for anything to be accepted into a standard are that there must be parity between data formats used and capabilities amongst competitors (in this case Nvidia, hence why from time to time there are joint roadmap discussions with software technology/tool providers, like Epic/Crytek) such parity wasn't there so it kept being delayed; so developers couldn't just assume say, a DirectX 9/10 part to be compliant. As for other specific differences, said pre-standard tesselation units lack documentation of how they differ generation from generation but it's probably down to data formats, vertex compression capabilities and the fact R7xx at least according to google still had to do tessellation in two passes as opposed to DirectX 11 one pass standard.
Even the worst case scenario (that they stayed Generation 2 and didn't go Generation 3/DirectX 11 Tesselation engine) is still better than X360, which is Generation 1. This generation nomenclature is not official but has been used in these threads before.
Current gen (R1xxx/HD 7xxx) tesselation units are considered Generation 3+.
lostinblue
Banned
That probably won't tell us much though. Unless they talk about it extensively or specifically.
Shin'en are technical prowess gods, but they're a very small team and thus the scope of their projects tends to be small and self contained (also very memory efficient, so that and extra RAM would help their case even if bound to the same tessellation unit as X360). Those are the circumstances even X360 tessellation unit is usable seeing the performance hit, and bloating of vertex data is more predictable in closed/not as demanding spaces; hence Viva Pinata and Project Milo using it on Xbox; both quite a bit self contained, challenge would be open world and simultaneously the place where modern tessellation units can help the most.
A game like Monolith's X(enothing) using it freely for game world geometry being what would set it completely apart (but we'll probably never know unless stated). Nintendo probably knew Tessellation was a very important part of the next gen feature-set definition though, so hopefully they borrowed the R8xx part as they did for Eyefinity. Even if it is the R7xx, it's a downer but still more usable than the X360 one.
ArchangelWest
Member
I've thought about this, and the fact that Devs have absolutely no problem with getting their games running on both screens, and the fact that there is not only never any latency between Game Pad screen and TV, but it's even been said that the Gamepad receives the images slightly faster.
Perhaps someone more understanding of the CPU and GPU partnership in the Wii U can ponder this idea:
Since the Wii U backwards compatibility is completely built into the chips themselves, then what if the game pad itself runs off of the Wii functionality, even while the chips are putting the Wii U game image on both screens?
Kind of like as though the Game pad is the first stop, so at the point where 480p resolution has been achieved, it passes through the hardware that broadcasts to the gamepad, yet then continues the pass through whatever cores are being used that cycle to give the on TV image?
I know I'm not explaining the thought too well, but I guess I'm saying that since the Wii U functionality has a physical point during each cycle where everything up until the next point is for all intents and purposes, Wii, then if the broadcast to the Game pad begins at that point, then not only is there no latency, but there's also no extra strain on the hardware from creating two images. Assuming that the transmission and reception tech is powerful enough. (Which it would seem to be due to the cost).
In any case, please excuse my rambling. Just a thought from a guy who doesn't understand this stuff anywhere near as well as you all.
*It also would fall in line with Nintendo's comments as to the machine "Doing exactly what you'd expect it to do".
Well, that would depend on the data being sent to the screen. I don't have hard numbers but if you're running a completely independent 3D scene on the tablet then it is using up whatever resources you wish it to. For instance, let's imagine a person is playing a 2 player fighting game on the TV and someone else is playing on the tablet on a different stage with different characters. Well then resources are nearly split 50/50 with differences due to resolution (so it wouldn't be an even split, it would depends on what resolution the devs allowed each screen to run at).
Now in the scenario where you're not using a gamepad, I would imagine keeping it totally black would give you nearly 100% for the main game. I know there is a pro controller but I am not sure if anyone is allowed to design Wii U software that does not use the gamepad at all.
lostinblue
Banned
First implied on a E3 2011 press release, it is also mentioned here ("Eyefinity-like multi-display tech for up to four SD video streams") and further detailed here in a pre-launch technical rundown article from game.watch.impress (known for publishing valuable japanese developer insight [1], [2], [3], etc).
Plus, it makes sense.
Shin Johnpv
Member
Yes it is not as advanced as the gen 3 tesselator in the DX11 cards. That said though you are miss remembering it as being disabled and unusable, specially not because it wasn't very advanced or wouldn't work with modern API's. The issue was with DX11 specifically. Which required a certain call(s) that AMD's gen 2 tesselator didn't support. How ever it could be used with other API's. Since Wii U wouldn't be using DX11 anyway, that's not an issue. You are correct that the base tesselator in the RV7xx is not as advanced as the one that would be in the PS4/720, but it's way more advanced than the one in the X360. However we don't know if part of the customization on the chip was to replace the RV7xx tesselator with a Gen 3 one, or to make other changes to the gen 2 one. Worst case scenario though is that it is exactly the same as the one on the RV7xx which is more advanced than the 360 one, and to try and claim it wouldn't be used, and that it would be treated the same as the 360's tesselator is just trolling and trying to bait fanboy comments.
Apophis2036
Banned
Good post and I agree with most of what you say, I just can't see third parties bothering with WiiU ports beyond Just Dance and yearly sports games. When CoD only sells 150k and AC3 only sells 100k on a system with an install base of 2.5 million then it's a warning sign for publishers.
I suppose it all comes back to a question I have asked a few times but has never been answered, how much does a port to WiiU actually cost.
Both of those games were highly hyped and people wanted them on day 1. Both were released on the Wii-U later than the other consoles. I wonder how much of an impact that really had.
Have we had any other AAA third party games come out since then that have come out on the same day on all 3 systems? I can't think of any, but I wonder if the attach rate would be any higher on the Wii-U then.
I just think a lot of Wii-U owners do own either a PS3 or a 360 too, so when those majorly hyped games came out even a week or two later, I think that hurt quite a bit.
But like you said, I think a lot of publishers will just look at the raw numbers and put up the orange to red warning flags.
lostinblue
Banned
TV sets can have more delay because they recieve said image from the source and then process it, they often have to convert it, scretch, take into account color mode and/or sharpening options plus gaming mode's and post processing, here (with the controller) nintendo controls the whole process and they invested/used top of the line tech to ensue that there was no lag.
You mean "if" the second screen is essentially being driven by Flipper/Hollywood GPU remains? Very unlikely as developers would have to be aware of it and code specifically for it (because of TEV/fixed function T&L). As for dedicated silicon to reduce the hit of running different stuff alongside (like extra ROP's) one can only speculate. I find it unlikely (the "Wii functionality" overhead, that is), I'd say at most they supercharged some areas of the GPU who would be more stressed out by doing dual screen in order to make it less noticeable.
Split screen is nothing major, hell, N64 could do it, 4 at a time at that. It has a hit, but it is nevertheless very doable.
Can't be. TV screen is most definitely still the main framebuffer when being used.
I doubt they do it in two independent threads as you're implying as I can imagine more disadvantages to it than advantages (but I'm not too certain that's what you're implying), nor would that be the only explanation for the controller getting the image before. It's a novel idea though, just counter productive.
Ah, no worries, I'm not the biggest tech head in here either.
- Status
Similar threads
- Poll
- Poll