Nintendo announce: Nintendo NX

[Shifty Geezer]

By and large they failed to get the Wii customers to migrate, and a significant part of the blame lies on the decision to include the tablet controller as opposed to refined motion controls.

I'll challenge that. Seems to me interest in Wii had died long before Wii U was a concept. It sold as a toy, a fun thing, with little longevity. Sales declined YOY after the first year or two on the market, where every other console grows until the next-gen. And taking one of the many images with data from the web...

Look where Wii software barely registers despite its explosive beginnings (fastest selling console ever in the UK). It's a tiny fraction of XB360 + PS3, both in 2013 and 2014. Wii owners had moved on, and HD Wii wouldn't have faired much better in the long term. There's no reason to think that Wii HD would have seen the longevity and software sales of the other mainstream consoles. Charts and stats for earlier in the lifecycle are similar, with Wii collapsing quickly. It's the poster child for the fad, and fads are rarely repeatable as it's often their novelty that makes them so popular, which can never be repeated.

 

[Entropy]

I didn't mean to say that it was the only reason Nintendo failed to bring Wii users over. A late launch, a price that was a bit too steep to encourage entry for casuals (where the controller contributed), a new control scheme that didn't really connect to the old one, customer base moving on as you point out, lack of really attractive Day1 software - there was a multitude of reasons, all being part of the problem.

But the controller contributed to at least three of those factors - the changed control scheme, the increased retail price, and since the WiiU was very obviously designed to a budget (the need to control costs even commented on at the time by Iwata), it may have contributed to not giving the console graphics performance that clearly distanced itself from the PS360.

In the context of the NX, the question is what conclusions Nintendo has drawn. Unfortunately I can't find the direct quote anymore, but here it is in another form:

Iwata went on to say there’s a need to “start over from zero” when putting out a “new game machine” and depending on the model change, there is a possibility of cutting off the “relationship with our customers”; therefore, Nintendo is “reflecting” on its past hardware releases and what a new system should offer to consumers.
“Switching platforms had [resulted] in a gap in the relationship with our customers,” he said (via Google translate).”Whether with Nintendo DS, 3DS, or what happened in the transition from Wii to Wii U
“I think it’s [something] to be reflected upon greatly.”

 

[Goodtwin]

But BC didn't dictate using an old/cheap 45nm process, or a very low power consumption limit or a very tiny CPU and GPU.
The only things BC demanded was using an old CPU architecture and perhaps the 32MB eDRAM to replace the 1T-SRAM. But It could've been a 6-core CPU based on the same Power 750 with 512KB cache each at 2GHz. And it could've been a GCN or VLIW4 GPU with 768 ALUs instead of the very old VLIW5 with 320 ALUs (according to some, it might even be only 160).

Regardless, I guess our disagreement stands on semantics
I would argue that going out of their way to provide backwards compatibility is more conservative than progressive. It means they're not confident enough that their new portfolio will cater enough interest from consumers, so they have to be on the defensive and keep supporting the old games.
I'm not saying BC is bad, not at all. Just that it's a conservative approach.

If you look at the constraints Nintendo's engineers were working with, the Wii U is actually a pretty sophisticated design. When you have to incorporate a controller that is certainly not inexpensive, maintain backwards compatibility, maintain a small size, and power efficiency must be there because of heat limitations, and then release for $299-349 without taking losses. From that perspective, im quite certain engineering the Wii U was more complicated than either the PS4 or X1. Wii U has only 12.8GB/s of memory bandwidth to its DDR3 memory, and according to developers, memory performance is good. That's a slick memory hierarchy that they came up with. Even the tri core CPU, a lot of people mock it for having roots going back to 1999, but that actually makes the PPC750 design that much more impressive, it is capable enough that various games on Wii U outclass its counterparts on PS3/360, and those CPU's were considered cutting edge when those console released. Did Nintendo inherently make it hard on themselves for features that the consumer cares very little about? Yea, probably so, but that's in the past, and now they need to develop a product that does have mass market appeal.

The success of the Wii probably clouded their judgment a bit, and were betting on a good chunk of that Wii market moving over to Wii U. That didn't happen, and why would it? The Wii customer base was primarily composed of people perfectly fine with the capabilities of the Wii, they have no interest in buying better hardware to play Wii Sports and Just Dance in HD. Not to mention the novelty of its appeal eventually wore off, and those customers moved on to other forms of entertainment. They didn't become gamers anymore than your Angry Birds audience did. Its a cash grab while it last, but its not a sustainable audience.

@Shifty Geezer

Your chart is a bit misleading. What were the big titles for Wii in the UK for 2013? Its last big title was in 2011, with Zelda SS. We are talking about a total neglect of software for the Wii in 2013. Imagine if Nintendo was still supporting it with a Wii Sports 3 and a new Mario Kart, im sure software sales would have looked a lot better.

 

[Rikimaru]

If you look at the constraints Nintendo's engineers were working with, the Wii U is actually a pretty sophisticated design. When you have to incorporate a controller that is certainly not inexpensive, maintain backwards compatibility, maintain a small size, and power efficiency must be there because of heat limitations, and then release for $299-349 without taking losses. From that perspective, im quite certain engineering the Wii U was more complicated than either the PS4 or X1. Wii U has only 12.8GB/s of memory bandwidth to its DDR3 memory, and according to developers, memory performance is good. That's a slick memory hierarchy that they came up with. Even the tri core CPU, a lot of people mock it for having roots going back to 1999, but that actually makes the PPC750 design that much more impressive, it is capable enough that various games on Wii U outclass its counterparts on PS3/360, and those CPU's were considered cutting edge when those console released. Did Nintendo inherently make it hard on themselves for features that the consumer cares very little about? Yea, probably so, but that's in the past, and now they need to develop a product that does have mass market appeal.

Wii U have been sold at loss. Memory system is "orthodox" as they said themselves (look Iwata Asks Wii U). PPC750 was indeed nice design, but SIMD is very lacking.

 

[Goodtwin]

Wii U have been sold at loss. Memory system is "orthodox" as they said themselves (look Iwata Asks Wii U). PPC750 was indeed nice design, but SIMD is very lacking.

Wii U hasn't been sold at a loss since early 2013, and it wasn't a big loss according to Reggie Filsaime. Their memory setup is hardly conventional, the PS4 has a conventional memory setup, Wii U relies very heavily on its edram for its GPU operations. It may be "orthodox" for Nintendo, who has used this type of memory setup since the Gamecube, but it hardly conventional compared to the like of PC and the PS4. The CPU also has very large L2 cache, so in theory, this should alleviate bandwidth needs to the DDR3 memory as well. The PPC750 design has miserable SIMD performance compared to the Xenon and Cell, but does in fact have its advantages in other areas. Those advantages weren't apparently able to offset the disadvantages, but exclusive games seem be able to avoid these pitfalls far more effectively than those porting games to the console.

 

[Exophase]

But It could've been a 6-core CPU based on the same Power 750 with 512KB cache each at 2GHz.

Not saying for sure it can't be done, but I've never seen a dual-issue OoO processor run at 2GHz with a tiny 4 stage pipeline. Most likely they would have needed a totally redesigned CPU to get significantly better performance (and you know, could have done with much needed improved SIMD)

Best thing to do, IMO, would have been to just have Wii's CPU integrated verbatim and use something totally different for Wii U games. The overhead wouldn't have been very high, a few mm^2 or so. They could have reused it some kind of peripheral processor if they really wanted to put it to use. I have a feeling that doing the same with the GPU would have been more efficient in the end too. Nintendo has gotten too clever for their own good in their attempts to integrate compatibility into new designs.

Their memory setup is hardly conventional, the PS4 has a conventional memory setup, Wii U relies very heavily on its edram for its GPU operations.

It's not conventional compared to a PC, but it's pretty standard for consoles. PS2, Gamecube, Wii, PSP, Nintendo DS, 3DS, XBox 360, and XBox One all had the same hierarchical design with a small(ish) fast integrated (usually DRAM) buffer(s) connected to the GPU and used for render targets (and usually texturing). Mobile GPUs that tile - and that's most of them - pretty much do the same thing too.

 

[AlNom]

Best thing to do, IMO, would have been to just have Wii's CPU integrated verbatim and use something totally different for Wii U games. The overhead wouldn't have been very high, a few mm^2 or so. They could have reused it some kind of peripheral processor if they really wanted to put it to use.

Sounds like it's still more expensive than bolting on two extra cores and reusing large portions of the SDK from the last decade (and no need to license another architecture, be it something else from IBM or ARM etc., more R&D). They would need some sort of fabric interconnect between the two architectures, no?

----

I'm not really clear on why they couldn't use the dead space on the CPU die to make all the L2 equal amongst the tri-core. Of course, one implication is that they're getting fewer defects in the eDRAM fab stage, but the chip being ~33mm^2 means wafer yields ought to be stupendous.

 

[Shifty Geezer]

Your chart is a bit misleading. What were the big titles for Wii in the UK for 2013? Its last big title was in 2011, with Zelda SS. We are talking about a total neglect of software for the Wii in 2013. Imagine if Nintendo was still supporting it with a Wii Sports 3 and a new Mario Kart, im sure software sales would have looked a lot better.

It was just one example. Searching for data, it mostly tops out around 2007 or 2010. I wanted something later. But 2007-2010 data shows similar, that interest in Wii exploded and collapsed, and the fact Nintendo weren't able to produce and sell software for their machine is still very relevant. If it was a viable platform with people buying software, companies would have been supplying software. That the UK sales for Wii software were so small shows there wasn't a viable market when the other consoles were going strong.

Wii's a great case study for originality in consoles. By and large, it doesn't work. And that's coming from me as a huge advocate for innovative controls! They just don't turn into new gaming paradigms and software falls back on old ideas, and platforms that can't execute those old ideas to the latest standards suffer. I can't recall a machine where that wasn't the case. Wii was the best champion for branching out and being original, but long term it wasn't great as a platform even if it was profitable for Nintendo. Kinect is another example of something new with considerable appeal, rapid sales, and no future.

By that token, Wii U was a good idea as it maintained the core experience, but execution was lacking.

NX seems a grass-roots re-evaluation. Games, standard games on standard hardware, just with more accessibility. Release on more, generic, platforms. Hardware can't be relied upon to be profitable so you need a more dependable if less exciting strategy.

 

[Goodtwin]

I still think looking at software numbers two years after the platform had been abandoned by not only third parties, but Nintendo themselves, skews the perception of Wii's falling popularity. Im not arguing that the public wasn't losing interest in the console, but I am arguing that had their been new quality releases on the console, software sales wouldn't suggest such a drastic plunge in consumer interest. There simply wasn't anything new to buy. Next year in 2016 when the PS3 and 360 have very few new releases, I would expect to see a huge decline in software sales on those platforms. Some of 360 and PS3's biggest software titles were released after Wii was abandoned. Again, 2011 was the last high profile release. If you look at Wii's yearly software sales, you will see it peaked in 2009, and then started to decline, but 2010-2011 were still solid years, but then in 2012, software sales were half what they were in 2011, and about a fourth of what they were at its peak in 2009. The bottom didn't fall out until the software stopped coming, that's my only point I guess. Your post suggest that the bottom fell out because of consumer interest going in the tank, and I don't feel that is really accurate. Declining consumer interest? Yes, but the bottom falling out was caused by no new software.

 

[Shifty Geezer]

There simply wasn't anything new to buy.

Why? That's the take-home point. XB360 and PS3 were just as old yet selling strong because people were releasing games on them because they were lively platforms.

 

[Goodtwin]

Why? That's the take-home point. XB360 and PS3 were just as old yet selling strong because people were releasing games on them because they were lively platforms.

My point is that if Nintendo hadnt stopped releasing games on Wii, consumer interest in the console wouldnt have fallen off the cliff. Your post suggest that consumer interest plummetted simply because consumers lost interest in the console, and I disagree with that. The main catalyst for consumer interest plumetting wasnt that the novelty of Wii taking a nose dive, but because software releases were non existent.

 

[Sonic]

My point is that if Nintendo hadnt stopped releasing games on Wii, consumer interest in the console wouldnt have fallen off the cliff. Your post suggest that consumer interest plummetted simply because consumers lost interest in the console, and I disagree with that. The main catalyst for consumer interest plumetting wasnt that the novelty of Wii taking a nose dive, but because software releases were non existent.

It was that the novelty of the experience had taken a nosedive and the market was moving in a direction the Wii simply couldn't exist in. In the beginning of the generation HDTV penetration was still low and the world entered a global recession. Nintendo benefit a lot from beikng the value choice and also I believe because HDTV penetration was still relatively low. PS360 became more affordable as time went on and so did HDTV's, Wii was left in the dust regardless of software support from Nintendo. And that's apparently held true this generation as well.

Nintendo is too conservative. While the engineering that has gone into Wii U may be something of a marvel given the restraints placed upon the engineers, that doesn't negate that it is a conservative design with gimped performance. I hope this is a mistake Nintendo doesn't repeat with its next system. Deliver on whatever brand new concept or gimmick they come up with and also deliver a more powerful console than PS4.

 

[Exophase]

My point is that if Nintendo hadnt stopped releasing games on Wii, consumer interest in the console wouldnt have fallen off the cliff. Your post suggest that consumer interest plummetted simply because consumers lost interest in the console, and I disagree with that. The main catalyst for consumer interest plumetting wasnt that the novelty of Wii taking a nose dive, but because software releases were non existent.

I guess that's one of the more subtle problems with having little more than first party support for a console: once they release a new console they'll want to throw all their effort behind supporting it at the expense of supporting the old one.

I wonder if that'll happen with this upcoming one.

 

[Shifty Geezer]

My point is that if Nintendo hadnt stopped releasing games on Wii, consumer interest in the console wouldnt have fallen off the cliff.

Why did people stop releasing games on Wii though? If it was a platform as healthy as the other two, the software should have kept on coming. There's a reason that the software stopped, and that ties in IMO with the games. I quote anecdotal evidence of people buying Wii's, enjoying them, and then putting them away in a cupboard with the board games to rarely be seen. Your counter argument is that Wii was a great, viable platform but Nintendo and other pubs just decided to stop making and selling software because they didn't want to make more money...

December 2010:

Nielsen finds the Xbox 360 users aged 13 years or older spend more time per week with their gaming system than PlayStation 3 or Wii users—4.9 hours per week, compared to 4.1 hours and 1.4 hours per week, respectively

Read more: http://www.digitaltrends.com/gaming...urs-but-wii-streams-more-video/#ixzz3Y1YmY12r

Here's another one
http://www.isfe.eu/sites/isfe.eu/files/video_gamers_in_europe_2010.pdf

European console owners playing their console for 6+ hours a week:
26% of Wii owners
39% of XB360 owners
40% PS3 owners

In Spring 2010, Wii was played less than the other consoles. (Interestingly, Nintendo posted a PR piece saying Wii was the most played console by taking the threshold at >2 hours per week). For light users it was the more popular machine, but for the core gamers, the people who spend money on software, it wasn't popular. Nintendo only made a killing off Wii because the hardware was so cheap and profitable. That was great business, but it's not a dependable strategy to apply to a new console or platform.

 

[Entropy]

I guess that's one of the more subtle problems with having little more than first party support for a console: once they release a new console they'll want to throw all their effort behind supporting it at the expense of supporting the old one.

I wonder if that'll happen with this upcoming one.

This is one of many reasons I believe the NX is a portable device. Sharing much between a game made for the 3DS and anything coming out in, say, autumn 2017 seems quite awkward. Between the WiiU and a portable device in the same timeframe, the mismatch is bound to be less, so ports should be doable and meaningful, and projects late in the life cycle of the WiiU could target both systems in parallel,
It's not a complete solution, but it would help. And it would help the WiiU as well, because it would remain a viable target for longer.

 

[Prophecy2k]

I don't understand how anyone can refer with admiration to the WiiU as a marvel of engineering. Especially by applying the contrived qualifier, "given the constraints placed upon the engineers".

If the constriants themselves, which were most importantly self-imposed, are fundamentally backwards, then there's nothing an engineer can produce that will qualify as an "engineering marvel" no matter how novel, unconventional or complex the design is.

As a chemical process engineer, if I design a new fangled reactor concept, for an application in an industrial chemical process to make ethanol from gasoline (for example), because I placed that artificial and poorly thought out constraint on myself. There won't be anyone who will call my design an "engineering marvel", when the application is forced to take a high value feedstock and convert it into a low value commodity product. Economically it's a non-starter.

A foundational aspect of engineering is first defining the problem you wish to solve. Nintendo decided that to solve their console gaming industry predicament and return to competitiveness in the market with a new console product, they would design a machine with a TDP budget so low that it would literally take the hand of God to achieve the success they were gunning for. They decided that HW performance was less important to consumers than BC and controller gimmicks, when the entire previous generation should have taught them the opposite.

The meteoric rise and crashing death of the Wii should have taught them that HW performance and by extension access to games (specifically thrid party titles) is more important to gamers than waggle and Nintendo's first party stable alone. Likewise, the complete lack of BC on the 360, total abandonment of it on the PS3 and increasingly lucrative sales of re-launched HD remasters should have taught them that BC in the box is just not important to gamers, and also that gamers are even prepared to rebuy old games on newer HW provided the price is right.

Nintendo decided to ignore all that and engineer the WiiU, essentially designing a console for mobile/tablet users, who struggle to pay their electricity bill, who also still have Wii games sat at home (without a Wii console) that also want to play them on a new console. It's the very definition of a solution for a problem that doesn't exist. Quite how anyone can call it an engineering marvel becuase it uses eDRAM I don't know...

 

[Entropy]

Christ, that was a rant and a half!
Nobody has called the WiiU an engineering marvel, it's all in your head. What some, me included, has said is that the design has integrity given its targets.
This thread isn't about the WiiU. That design is only interesting insofar as it can serve to predict what Nintendo will decide to do with the NX.
Now, I happen to think that you can hypothesize from the WiiU, but mostly about price points and such. What conclusions Nintendo has drawn from the market fate of the WiiU is anybodys guess, other than that they will try to avoid the same thing happening again. What that means for a dedicated gaming product that we don't really know when it will be released, or even if it is portable or stationary, well....

 

[Shifty Geezer]

I don't understand how anyone can refer with admiration to the WiiU as a marvel of engineering. Especially by applying the contrived qualifier, "given the constraints placed upon the engineers".

Actually, engineering can be impressive when it solves a problem, even if the problem is self-inflicted (save when the problem is self-inflected by poor engineering). The management asks for the engineers to solve something, and solving that effectively is where the engineers can be marvels or not. "We want a five wheeled car with a wheel in the middle of the four corners, with three wheel steering and four wheel drive." If the engineers can pull that off well, it's a good job, even though the product is possibly moronic.

Not that I think Wii U is well engineered, but I think it's right to separate the engineering solutions from the product designs where the designs aren't a product of the engineers.

 

[Deleted member 13524]

My point is that if Nintendo hadnt stopped releasing games on Wii, consumer interest in the console wouldnt have fallen off the cliff.

It were Nintendo's first party sales that drove them away from releasing more titles on the Wii.

Twilight Princess was a release title in 2006 for the Wii. Initial userbase was zero, the game wasn't even bundled with the console during release AFAIR. Moreover, the game was also released on the Gamecube and had almost the exact same visuals, so some people bought it for the older console. Twilight Princess sold around 6 million copies for the Wii alone and another 1.3 million for the Gamecube.

Fast forward to 2011 and Skyward Sword gets released exclusively on the Wii. Review scores are also off-the-charts, one of the top games in Metacritic and this time the game counts with a userbase of 95 million consoles. It sold 3.4 million copies.
It means that, by November 2011, less than 4% of Wii owners were buying games for the console, and not even Zelda swayed their minds.

And it's not like Nintendo gave people a "drought" of first party titles before that. Metroid: Other M, Mario Galaxy 2 and Donkey Kong Country Returns had been released in 2010.
Nintendo did all they could to keep the platform alive. It's just that their first and second-parties aren't enough to keep a console going.

 

[Syferz]

This went off topic fast.

No one came here to talk about how Nintendo abandoned the Wii before it's audience did. (Yes, Nintendo slow trickled titles to the Wii while talking about 3DS and Wii U during 2011, the success of 2010 on Wii and the lack of software in 2011 is quite black and white, Nintendo had moved on and so did the gamers) If Nintendo learned anything from the Wii generation, it was that they could be more successful in a niche in half the time than their competitors in the mainstream who have yet to match Wii's numbers. Lots of people claim lot software attach rates but Wii's software attach rate is nearly 9:1 which is the 2nd highest attach rate of any Nintendo device at ~900 million software titles sold, only loosing to the much less successful Gamecube with ~9.5:1 attach rate.

No one came here to talk about the Wii U and how well/poorly it was put together, the choice of going with an MCM however does mean that BC is as simple as adding Espresso as a co processor much the same way they do with their handhelds. 45nm process for the CPU was chosen because Wii's CPU was shrunk to 45nm at the time, and 40nm for the GPU was chosen because Resna (sp?) had to produce the EDRAM inside the GPU die, and this was their smallest process node. Nintendo shooting for a system under 40watts is where they really limited themselves, Iwata had once said that it was really hard to balance performance with their goals and that at one time he thought it wouldn't be enough, low performance was never their goal, but it was exactly where they headed thanks to those limitations.

NX has completely different goals, from what we know: (speculation in parentheses)
1. Will adequately absorb Wii U architecture. (likely MCM with a coprocessor)
2. Moving their audience to an account based system and away from individual hardware devices. (Shared library)
3. Dedicated software engineers were working on the NX no later than Feb. 2014.
4. AMD is making an APU for a gaming console that will be announced in 2016. (Likely NX)
5. Nintendo is thinking about expanding their device offerings to 3 or more devices, starting with the "NX"