Wii U hardware discussion and investigation *rename

While I referred to iSuppli numbers, they are mostly without consequence to investors, who care about profitability. Ultimately, the Bill Of Materials is mostly useless to judge overall Return On Investment. My guess would be that Reggies statement concerned the "initial batch", or some interval in time specific to internal accounting. And if so, it would be eminently reasonable to include costs that are specific to the launch of the device. (R&D, software tool production, middleware procurement and so on, I'd assume to lie outside such accounting, but is still relevant to overall ROI.)

We don't know how they account for costs associated with securing manufacturing contracts, producing commercials and other marketing material, including launch associated air time and ad space, money hatting, warehousing and distribution, additional software development costs and deals necessary for added services, costs associated with setting up servers, design of retail boxes and printed material, and localization of said material,.... the list goes on and on and on.

It makes sense to have the launch associated costs accounted for in the launch window as well, even though the benefits extend over the full lifetime of the console. After all, the financial burden is now. Also, it makes sense to associate them with the relevant product.

Even when we get an iSuppli breakdown of the WiiU, it bears remembering that the BOM in no way, shape or form constitute the total cost of offering the product to the public. So many times I've seen the BOM naively interpreted as the total cost and indignation over the indecent margins a manufacturer has on a device. Let's try to do better here.

 

From what I see, Wii U just needs a little push to get that better ports. If they can double the CPU speed and use faster memory, it should be enough. Of course I don't know how viable it is, but their power budget is definitely too low to have a hardware that have a clear gap between it and the last gen (PS360). I felt that the whole thing is not just limited to the money budget but also the power budget. Why are they limiting themselves to such a small power draw? Is there a logical answer to this? The current Wii U power draw is practically in laptop territory. Is there a financial benefits for going "green"?

 

This is simply untrue. I believe you are trying to up play any advantage the WiiU may have and at the same time downplaying most rumors about the actual next gen systems. There's very little room to even suggest "And following the recent rumors about PS4 and Xbox3 ... they honestly don't seem much more powerful than Wii U" in an honest manner based upon rumors and speculation. The WiiU will not be the PS2 of the coming generation as it will be the Wii of the new generation, most likely without the gargantuan sales numbers.

Now it may be that some devs make miracles happen with this hardware, but that really doesn't matter as the new Sony/MS consoles will still crush the system in terms of power and performance in just about every single way. And whatever improvements we see for the second save of WiiU games will most likely be crushed by what we see of first generation next gen games. And once the second wave of games hit in 2014, the WiiU will be left far behind in a graphical sense.

 

Perhaps the CPU couldn't scale any further, even with additional power. The GC/Wii CPU were supposed to have incredibly short pipelines. Even with the 45nm shrink it might not go much beyond 1.25 gHz ...

A couple of Bobcat cores at 1.65 gHz would probably have been much more capable than the 3 cores Nintendo got and they could have been integrated into the GPU die from day 1. They could even have added a 90nm Wii CPU to the package for BC (they're already making them for the Wii and they must cost pennies now) and yanked them out when it was time to kill BC.

I still really, really want to know about edram bandwidth. I still have a feeling its around (or maybe less) than the PS3s vram and I wouldn't even bet that the CPU has equal, unfettered access to it (and I don't think BC would require that either).

 

So basically the low power draw (at least as low as the end product) isn't really intentional but they just can't extract more performance from the silicon (at a reasonable yield that still within their money budget)?

 

I can't know that for sure, but I think they'll be at their target point for yields and that nudging the clockspeed up would start to need nasty amounts of additional power, where the cost throughout the system (power supply including all the stuff on the board, cooling, case size, fan noise) wouldn't justify the increasingly tiny amounts of additional performance.

Even if you're prepared to accept poor yields and huge power consumption there's only so far an architecture can scale upwards in terms of speed (and downwards in terms of cost/power effectiveness). I'd guess the Wii U is toward the upper limit of clockspeed, and that any gains from pumping lots of extra power into the chip would be both very small and relatively expensive.

There are people on B3D that are really knowledgeable on this kind of stuff, hopefully one of them will chip in with their opinion...?

 

I have a question about the GPU. Based on the teardowns, it's relatively large, right? I heard it's the size of an RV740 if we assume 40nm. Despite this, it barely offers RV730 performance. Does anyone have a clue why it's so big? Could it really be that that heavily customized, or was Nintendo really dumb enough to put in a 55nm GPU?

 

They'll be including more than just BOM when talking about the cost of a unit.

The PS360 is old as hell and gone through several cost drops in that time. The Wii U, whilst not uber powerful is still modern. There's still cost in the designing of the system and manufacture of the custom parts. They're not off the shelf. As Reggie said it only takes one game to become profitable and really we don' t know if that means it takes $1 from that game sale to be profitable or $30.

I agree it's perplexing, but we can't assume that because we can get a shitty tablet and an old PS3 for the same price that Wii U must be cheap as buggery to make.

Well they have said they are. Why should we assume they are lying?

 

Keep in mind the BOM of those consoles were significantly above their retail prices. They've dropped significantly through redesigns and die shrinks that make both of them still much larger than the WiiU. And as new as the wiiu is, there really isn't anything state of the art about the hardware. There is definitely some Hollywood accounting going on to reach the statement that the hardware is sold at a loss, that or they are giving retail significant margins.

 

it depends what do you mean by lying, if you include in your definition : misleading statements (as others said some mumbo jumbo accounting definition of manufacturing costs per console) than yes they are most probably lying, maybe they are losing money for WiiU basic version at 250$ in the US, but in europe for example there is no way they are loosing money on every unit bundled with one game at 350 euros. and again it never happened to intendo to loose money with their previous consoles even though they were relatively powerful compared to their competitors (N64, Game Cube), why this would be any different for an underpowered old hardware console a la WiiU ?

 

Yeah, I think it's a valid point to look at different models and different regions. According to the webs:
http://www.gamesindustry.biz/articl...akes-one-game-sale-to-make-a-wii-u-profitable

Doesn't specify which model, or what type of software. So the peasant model with 32KB of user accessible flash at $300, when sold with a third party game for $10 from the e-shop (with Nintendo probably getting about 1/3 of that) will presumably be profitable.

Of course, no-one is buying the peasant model because it represents the worst value ever. Prices are higher in the UK, so Nintendo are probably in profit on everything. Trade prices here are reported to be £205 and £246, meaning that even the basic model is getting Nintendo ~ $330.

 

Lets be fair and realize that Nintendo did not make a console to replace the 360 or PS3. They made a console to replace the Wii. So, the WiiU is a console easily outperforming its predecessor.

We may want to compare the WiiU to the 360/PS3, but obviously this is not Nintendo's main focus. Their focus is making a console for presenting their games in high definition, making use of the gamepad. This did not require them to produce a console to outperform the current HD twins. Maybe the WiiU will eventually show that it can outperform the HD twins and hold its own against the new consoles.

And lets also not forget Nintendo is making consoles with 5 year life-spans. While Sony and MS decided for a longer lasting generation. Who knows what they are planning now.

At any rate, MS and Sony, have to develop consoles to outperform their current consoles. The performance leap has to be noticeably better than what is currently being offered. This probably means manufacturing an expensive product, especially considering they will include an alternative to the gamepad, i.e. Kinect2 in the box.

 

But what are you basing this on? Quick, budget ports put together by "C" teams, or outsourced to untested development teams? eShop games? Because I haven't seen any AAA game made exclusively on the WiiU so far. Especially not first party. Pikmin3 is being made by a small team, so its not AAA. ZombiU is not AAA, it is a small budget game.

I think March will probably be the first time we are going to see big WiiU games being presented.

 

But thats not what they were trying to achieve. They wanted to make the package small. Adding the Wii components wouldn't allow that. The smaller the package, the less heat, the more performance they could get out of the console.

What I dont get is that porting games between consoles and PC's have always been a problem. Yet for the WiiU everyone expects it to magically take code from a differently designed machine and run it wonderfully. And when that doesn't happen it must be because the hardware sucks. I dont understand this jumping to conclusions based on launch games. When 360 games dont run well on PC's is it because the PC's are inferior?

 

I think a difference is that we don't have any indication that it is difficult to program for. A Wii core should be really well documented by now. And DX10 GPUs. The console was billed as an easy porting machine from the start.

 

On the one side this is not the PS3 we are talking about. There are no paradigm shift or anything on how you treat your code or anything like that.

On the other hand, I do agree that we should not jump to premature conclusions either. It most likely has some particular weaknesses that could be overcome if the software was better targeted to the platform.

I feel the general consensus though is that one should expect the WiiU to be powerful enough that even less elegant code should be able to run well just because of brute force...

 

Wii U Gamepad panel type

Have we got a definite answer now on what type of panel the gamepad uses?

I have one, the blacks are very poor (not surprising) has a decent amount of backlight bleed, but I'm finding it hard finding out what panel it is, such as IPS or TN.

 

TN is easily given away, tilt it and if from one direction it becomes dark and weird with colors totally wrong, it's a TN (same thing as looking at a typical desktop or laptop LCD from below)

It's very reasonable to have a TN on a handheld controller.

 

I understand that, but with a game like NSMB U, which is very colourful, tilting the screen doesn't show the colours becoming washed out or like a negative.

But something that is dark, such as COD BLOPS II's title screen which is mostly black, the colours wash out, but I believe it's a backlight issue (the backlight bleeds from an angle).

Just took some pics, apologies for the quality.

 

It's definitely not (regular) TN, the colors and contrast don't go all buggo when tilting the controller like they do in cheap laptop displays if that'd been the case. Colors stay quite solid overall, even at strong angles of tilt.

If you tilt and rotate the controller 45 degrees however there is a pronounced contrast shift. Darker greys go light, etc. Not that this is likely to occur during regular gameplay though as you need to tilt it a lot for this to show up. The screen is barely viewable as it is at such angles and no game would rely on such extremes of motion.

The panel might possibly be some form of higher-end VA, or possibly a simpler IPS variety. The contrast shift doesn't appear on my iPhone, which uses a high-end IPS panel of course.