Nintendo announce: Nintendo NX

Status
Not open for further replies.
I still expect end of year release with hopefully (for Nintendo) the next Zelda both releasing on WiiU and NX, maybe a bit later on NX, similar to what they did with Twilight princess on gamecube and Wii.

And it won't be 4K@60fps* lol on NX. Best case scenario: (gamecube -> Wii) == (WiiU -> NX). Worst case: WiiU == NX

* PS4 / XB1 level specs won't be possible on a tablet-ish device. :nope:
 
Even on 14 nm, WiiU will probably be too power hungry to go fanless and have good battery life. The architecutre(s) are simply too old and not designed with that aim in mind.

I really hope that Nintendo don't recycle the doomed WiiU hardware. There's the perennial joke about Nintendo and third parties, but if Nintendo want a home console in the traditional sense they've got to support modern software and that means coming in around Xbox One. But the question is, as always: is this what they actually want, and do Nintendo actually care ... this time.
 
I do think ARM is very likely for the CPU. There are just so many things going against an IBM design (much more than 5 years ago). It does seem that Kimishima is also prepping people for a lack of BC when he talks about a clean break from the Wii/Wii U line. Still, it makes me wonder, because the WSJ article (a stationary unit w/ at least one mobile unit) sounds alot like a better version of Wii U. Right now, I think NX will come in at around 1 TFLOP. It will use a more modern AMD SoC w/ ARM for the CPU. But from what we've read, I don't know if their ideas will be original and exciting enough to make a big impact on the market.
 
I suppose we should actually be considering ... Nvidia.

The have a power efficient GPU architecture that can scale from fanless handheld up to industry dominating 980s and Titans. They also have a pretty fast, power efficient ARM setup that would be perfect in a handheld and probably scale up to be competitive with Jaguar. And they can deliver all this on 28nm or 20 nm, today, using LPDDR4 or GDDR5.

And Nintendo haven't been Nvidia'd yet.
 
For the sake of argument though, what would it take to be 20 times faster in practice?

If there are 160 VLIW5 shaders at 550MHz in the Wii U, then it does a theoretical 176 GFLOP/s. 20 times that would be 3.52 TFLOP/s. A R9 380 (Tonga Pro with 28 active CUs at 970MHz) would do just that, but with a probably much higher real-world performance, since GCN is much more effective at keeping ALU utilization up.

Tonga is a big chip measuring 360mm^2 at 28nm, but using FinFet it would probably go down to 160mm^2 or so.
If they used FinFet, Nintendo could do 20x the Wii U's GPU performance with a 200mm^2 APU, given they didn't spend a big chunk of transistors on embedded RAM.
 
It looks like all console makers have had a bad experience with nVidia so far. Maybe they're burned among that industry.
 
If there are 160 VLIW5 shaders at 550MHz in the Wii U, then it does a theoretical 176 GFLOP/s. 20 times that would be 3.52 TFLOP/s. A R9 380 (Tonga Pro with 28 active CUs at 970MHz) would do just that, but with a probably much higher real-world performance, since GCN is much more effective at keeping ALU utilization up.

Tonga is a big chip measuring 360mm^2 at 28nm, but using FinFet it would probably go down to 160mm^2 or so.
If they used FinFet, Nintendo could do 20x the Wii U's GPU performance with a 200mm^2 APU, given they didn't spend a big chunk of transistors on embedded RAM.

Yeah, the chips to do it in pure FLOPS terms are definitely out there, but it's the whole "efficiency" thing that's interesting. Roughly speaking, it seems like GCN is something like a quarter more efficient "per flop" than VLIW5 (as cumbersome and vague as that sounds!). 200 mm^2 for the APU sounds reasonable, especially given a 256-bit bus and perhaps four faster GPU cores.

If Nintendo were to go for 14 nm my (dodgy) guts tell me that they might aim smaller than that though...
 
If Nintendo were to go for 14 nm my (dodgy) guts tell me that they might aim smaller than that though...

That's true. If they go 14nm there's a chance they'll go for a lot less than that.


The other day I went to see how big the previous chips from Nintendo were, and it turns out the Wii U actually has the largest sum of CPU+GPU die areas since the Nintendo 64.
The Gamecube had a 120mm^2 GPU + 44mm^2 CPU, but it was made using 180nm which was the best process available at the time.
Then the Wii was just laughably tiny with a 16mm^2 CPU and a 72m^2 GPU (basically a Gamecube shrunk to 90nm).
The Wii U was actually large-ish with a 33mm^2 CPU and 156mm^2 GPU.

So Gamecube was small-ish but it used the latest process node at the time. The Wii was all lol I don't care I'm just gonna be cheap! but it also used the latest process available at the time.
Only with the Wii U did Nintendo get greedy with the fab technology, and decided to go with a ~5 year-old process by the time the console shipped.
 
Do we really think 3rd parties would be lining up to develop launch titles once it's announced, such that Nintendo need to wait 2 years after announcing before they can launch? Whatever it is, 3rd parties will wait and see save for the cheap toe-dipping options. So Nintendo will have to provude the library that sustains NX, and that could/should be progressed now such that it's ready six months after the announcement (or however long).

Depends on what third party developers we are talking about. If the NX has a portable form factor from day one, then we can assume 3DS developers will be there from day one. Not to mention a wide variety of first party exclusives from Nintendo. If NX consolidates development into one ecosystem instead of being split between console and portable, then I can see Nintendo having a really strong lineup of software ready to go. Nintendo has also done well to get Indies on board with the 3DS and Wii U, so they will be there on day one as well.

As for the big third party publishers like Ubisoft and EA, it will likely come down to ease of porting. If its easy to port their games, they will do it for the easy money, but if its difficult they will most likely take the wait and see approach. Its tough to say if they will want to compete on the platform unless porting their games is a viable option. The 3DS has sold well, but the big third party publishers don't really support that platform.
 
While I'm sure the Gamepad factored into the cost more so than an average controller, I'd also think the processor(s) probably ended up costing them a bunch. Despite the relatively lean amount of silicon, you've got 3 different companies (Renesas, AMD, IBM) and the licensing fees involved with that. IBM gave them a custom chip using an SOI process, which adds to cost. You had the Renesas eDRAM and all the costs and custom process node to go along with that. Also, the Xbox 360 is sitting at $199 even now. The last generation of consoles simply weren't able to be reduced in price as much as previous generations.

Let's say that Reggie meant the basic console when he was talking about having to sell a first party game along with the console to break even, that's still over $300 (taking into account a retailers cut) that they were looking to make back, I just don't see how WiiU could have possibly cost them that much to manufacture esp when they didn't even include basics like an Ethernet port... Even being conservative I would say the Gamepad cost them at least $100 to manufacture at launch and for much of 2013.

Didn't one of the guys from chipworks in the gaf thread estimate the main MCM to cost ~$100 ? If we are taking that as fact then I really don't believe that the WiiU's eDRAM, DDR3 RAM, WIFI chip, ARM CPU, pitiful storage, disk drive, tiny fan and case cost anywhere in the region of $170 to manufacture in late 2012 if you're going on the notion of a $30 to manufacture Gamepad for instance. It makes much more sense to estimate the main MCM to cost ~ $100, the other components combined to cost ~$100 and the Gamepad to cost ~$100 giving us the $300 Nintendo needed to break even (it could well have been closer to $350 they needed to break even if Reggie meant the Deluxe pack).

I also feel the MSRP of the console backs up the case for a $100 Gamepad. You could argue that they aren't cutting the cost of the console now because it's a lost cause but I think we would all agree that if Nintendo could have price cut WiiU to $249 or even $199 in late 2013 or mid to late 2014 you can bet they would have done it esp with the likes of 3D World, MK8 and Smash launching during those times. I believe they simply couldn't do it without incurring massive financial losses caused by the Gamepad's high manufacturing costs. The main MCM aswell as all the other components (outside of the eDRAM because of Sony's factory purchase) would have drastically dropped in manufacturing costs by that time aswell meaning they could have cut to $249 or $199 if the Gamepad really did only cost $30 to manufacture.

In hindsight they should have removed the Gamepad from the console, bundled a pro controller as standard and price dropped the console to $199 and completely re branded the console in time for MK8 launching, it wouldn't have turned the consoles fortunes around but it would have been sitting at closer to 20 million sales now instead of 10 million imho.

However powerful NX is I hope they follow Sony's lead of using cheap, standard off the shelf components that can be price dropped a couple of years into it's life because the sooner they hit the $199 - $149 mark the more appealing the console becomes to their main demographics - families, kids, core gamers as a second 'exclusives only' console. Releasing at $300 or more will see them again only appeal to the same hardcore Nintendo fans that bought WiiU.
 
Do we really think 3rd parties would be lining up to develop launch titles once it's announced, such that Nintendo need to wait 2 years after announcing before they can launch? Whatever it is, 3rd parties will wait and see save for the cheap toe-dipping options. So Nintendo will have to provide the library that sustains NX, and that could/should be progressed now such that it's ready six months after the announcement (or however long).
Initially I agreed with you, but I just thought of something
Nintendo can always offer a certain amount of financial reimbursement on the cost of porting/shipping/buying floorspace for various titles it wants on its system if they don't sell. Might cost a few million for a big title, and cumulatively a hundred million over the initial launch period when the consumer base is small.
But Nintendo is sitting on 10 billion in cash reserves so its probably a long term net benefit to reestablish trust in consumers that NX is a 3rd party rich platform.

I doubt I'm the 1st person to think of using about 1-2% of their strategic cash reserves.
 
Last edited:
Off topic, but the gamepad is a selling point of the Wii U to me, Nintendo's games and the ability to couch coop are also, obviously major points.
The gamepad, I'll never stress it enough, is like a VR helmet w/o cutting yourself from reality, minus 3D though, it provides similar functionality with fewer drawbacks. (But I know kids want to get immersed in another world, I would have liked it when I was a kid, but reality is much more important to me now. Also sight is only one sense, if everything else don't match it I'm dubious it will improve much at all, hence HTC Vive being more likely to be a good experience/the future in that regard).

Anyway back on topic, I have no idea regarding third parties, it's always good to have more options, but it's interesting if either : the titles are quality wise on par with other versions, or the titles are exclusive.
 
The tablet controller is a selling point to me too. However, I'd rather have a decently performing console and get the tablet as a peripheral than to get a crappy console with tablet included that will be overlooked by third parties for being crappy.
 
Nintendo did a terrible job of really showcasing how the Gamepad improves the experience. Nintendo Land showed how it can be cool with mini games, but other than that Nintendo didn't really have any core gaming software that made great use of its features. At launch, the best game to really make good use was Zombi U, and true survival horror games aren't really a hot genre these days. Some games like Mass Effect 3 and Deus Ex made great use of it for menu's, but this was more of a convenience feature, and not a game changing feature. For some people, offscreen play is a big deal, but for most people its not. I like the Gamepad, but I have to believe that without the Gamepad, they could have sold the Wii U as a Wii HD that came with a Pro Controller, probably launched at $199 or $249 at the most, and would be $149 by now and probably selling a lot better than it is.
 
Wii U Pad is atrocious. Bulky and heavy. Not suitable for a long comfortable play.
Screen is terrible. Video compression too. 3hrs life.

Brings nothing to game experience.
I play most games on Pro Controller which is much more comfortable. I've even removed battery from Wii U Pad.
 
I suppose we should actually be considering ... Nvidia.

The have a power efficient GPU architecture that can scale from fanless handheld up to industry dominating 980s and Titans. They also have a pretty fast, power efficient ARM setup that would be perfect in a handheld and probably scale up to be competitive with Jaguar. And they can deliver all this on 28nm or 20 nm, today, using LPDDR4 or GDDR5.

And Nintendo haven't been Nvidia'd yet.


One way or another I'm kind of almost expecting Nvidia in the next round. Whether through their own console or building somebody else's.

It just makes too much sense IMO, not like they've got a whole lot else going on in terms of expansion (with their foray into mobile being relatively unsuccessful), and if they've got time/money for vanity projects like Shield they've got time for a console. They can partner easily with ARM (or design their own with arm license of course) on the CPU side that's no issue.

Doubt it would be NX though, although I guess it could.
 
Status
Not open for further replies.
Back
Top