Nintendo DX GPU?

[RussianSensation]

So many people keep saying that Nintendo needs to release a console way more powerful than XB1/PS4 but I disagree for the following reasons:

1) Let's say just throw it out there and assume NX is 3-4X faster than PS4/XB1. Would that matter? No. The reason it wouldn't matter is that by the time the NX releases in 2016, XB1/PS4 will have an install base of at least 50 million units. The NX will be a new kid on the block. Why would 3rd party developers spend extra effort to take advantage of much more powerful Nintendo hardware when there will be so few NX owners in the first 2-3 years of the console's life? After all, even hardware 3-4X more powerful isn't just going to sell the console when "most of your friends have XB1/PS4 or both and console games like playing with their friends." It also means the NX will have to contend with by then a huge library of games to choose from among XB1/PS4 consoles. More powerful hardware also has to go hand-in-hand with Nintendo actually getting more 3rd party support, which in itself is doubtful given the track record of Nintendo's 3rd party support since N64. If they can't secure a lot of 3rd parties, what's the point of all that extra horsepower? This is because Nintendo itself isn't the type of company that spends hundred million dollars making the most gorgeous game and honestly I doubt their team even has the technical ability to do so. We wouldn't suddenly see the next Zelda, Mario, F-Zero game blow the doors off Uncharted 4, Rise of the Tomb Raider, Gran Turismo 7 because Nintendo isn't a firm that knows how to make the best looking 1st party games either. In simplest terms, given how underpowered the Wii U is, even a GPU at XB1 level is more than enough for Nintendo's game designers to make much better looking games. They don't need the power of the Nano or 980Ti in there.

2) Adding a GPU far more powerful than XB1/PS4 would mean a more expensive console since Nintendo is reluctant to lose $ on hardware. By end of 2016, it's possible there will be even more price cuts on PS4/XB1. Nintendo cannot afford to launch a $450-500 console with 3-4X the power of PS4 when by 2016 it'll be easy to pick up XB1/PS4 for $299. This is another reason going with a powerful Nintendo GPU is too risky.

Another way to look at it is this -- if the next Nintendo console is much more powerful than the existing consoles, by the time we would start to see some games taking advantage of that hardware, it'll be 2-3 years and by 2018-2019 we could be looking at XB2/PS5 which could be 6-8X more powerful than XB1/PS4 and again trounce the 2016 Nintendo console.

For that reason, Nintendo needs to focus on offering something unique or focus on low price. The strategy of a powerful Nintendo console would have been far more successful if they launched it in 2013 but not mid-cycle in the generation.

I would honestly be surprised if the next Nintendo console is actually more powerful than the PS4.

 

[RussianSensation]

I'm betting something custom on the cpu side faster than both the ps4 and xbox 1.
And a power vr graphics that are better than the ps4.

People keep talking smack about the graphics chip in PS4 without realizing just how difficult it would be to significantly outperform it, especially in APU form. Also, your comment about PowerVR graphics beating PS4 is 3-4 years away. The fastest smartphone SoC's graphics is generations behind HD7850/R7 370 and it won't have any hope of catching up in 2016 either.

R7 370 = 975mhz clocks 1024 SPs, 32 ROPs, 64 TMUs, 256-bit bus @ 5600mhz GDDR5, based on GCN 1.0 which means just 2 ACE engines

vs. PS4

800mhz clocks 1152 SPs, 32 ROPs, 72 TMUs, 256-bit bus @ 5500mhz GDDR5, based on custom GCN architecture, with 8 ACEs that support up to 64 compute command ques, volatile tags for L2 cache, custom 20GB/sec data path between the CPU and GPU (that's faster speed than PCIe 3.0 x16).
http://www.gamasutra.com/view/feature/191007/inside_the_playstation_4_with_mark_.php?print=1

That means R7 370 is a very good estimate at minimum as to how fast the GPU in the PS4 is at its full potential.

AMD's current stack vs. R7 370:

285 is 39% faster
280X is 50% faster
R9 290 is 90% faster

And now the power usage.

280X would use at least 225W as 925mhz 7970 uses 201W.
290 uses 263W for reference design.
285 is less power efficient than the HD7950, which uses 149W.

On 28nm with conventional GDDR5 256-bit bus, a reasonable scenario for Nintendo is a 1280 shader APU which is barely better than PS4.

Now an astute AT forum member would have noticed that an HD7850/R7 265 or R7 370 all outperform a GTX750Ti 2GB, but yet in the real world, a Core i3/FX4300 + GTX750Ti would easily outperform the PS4 in most games. Why is that? That's because 1.6-1.75Ghz Jaguar CPU in the current consoles is a huge bottleneck for even an HD7850/R7 265/R7 370 style GPU.

That means the ONLY way for Nintendo to measurably increase GPU performance beyond PS4 is to incorporate a way faster CPU, which will only raise the total cost of the APU and power usage. Unless Nintendo decides to lose $ on hardware, it's a very slim chance for their console to have a much faster GPU than the 1280 shader GCN 1.2/1.3 because going beyond that is only going to worsen the CPU bottleneck.

Realistically speaking, all Nintendo needs is a 896-1152 shader GPU and 8 Jaguar cores clocked at 2.0Ghz and the rest they should focus on other aspects like online gaming, etc.

 

[stahlhart]

Thread cleaned. Please refer back to the OP, and stay on topic.
-- stahlhart

 

[el etro]

I would like to see at least Fury Nano performance on the console, but is a hard thing to achieve even on 14/16FF. I want Nintendo spending all it can for bleeding edge tech on the new console. Maybe Hawaii performance GPU, plus 4x ARM A72 cores, plus 2x ARM A53 cores, with no handheld attached on it.

 

[Zstream]

Heck a 2.4 ghz jaguar can be achieved without much power increase. I'd like to see 4x2.4 vs 8x1.6ghz benchmark.

 

[Techhog]

I would like to see at least Fury Nano performance on the console, but is a hard thing to achieve even on 14/16FF. I want Nintendo spending all it can for bleeding edge tech on the new console. Maybe Hawaii performance GPU, plus 4x ARM A72 cores, plus 2x ARM A53 cores, with no handheld attached on it.

That would just be throwing money out of the window. It wouldn't be a real selling point, and the console would either be priced too high or would bankrupt Nintendo. Sorry, but you need to temper your expectations. They're going to aim for $250 or lower, and it'll be slower than the XBox One. These are practically givens. It'll be fast enough to play the handheld's games in 1080p, and that's it.

 

[EarthwormJim]

People keep talking smack about the graphics chip in PS4 without realizing just how difficult it would be to significantly outperform it, especially in APU form. Also, your comment about PowerVR graphics beating PS4 is 3-4 years away. The fastest smartphone SoC's graphics is generations behind HD7850/R7 370 and it won't have any hope of catching up in 2016 either.

R7 370 = 975mhz clocks 1024 SPs, 32 ROPs, 64 TMUs, 256-bit bus @ 5600mhz GDDR5, based on GCN 1.0 which means just 2 ACE engines

vs. PS4

800mhz clocks 1152 SPs, 32 ROPs, 72 TMUs, 256-bit bus @ 5500mhz GDDR5, based on custom GCN architecture, with 8 ACEs that support up to 64 compute command ques, volatile tags for L2 cache, custom 20GB/sec data path between the CPU and GPU (that's faster speed than PCIe 3.0 x16).
http://www.gamasutra.com/view/feature/191007/inside_the_playstation_4_with_mark_.php?print=1

That means R7 370 is a very good estimate at minimum as to how fast the GPU in the PS4 is at its full potential.

AMD's current stack vs. R7 370:

285 is 39% faster
280X is 50% faster
R9 290 is 90% faster


And now the power usage.

280X would use at least 225W as 925mhz 7970 uses 201W.
290 uses 263W for reference design.
285 is less power efficient than the HD7950, which uses 149W.

On 28nm with conventional GDDR5 256-bit bus, a reasonable scenario for Nintendo is a 1280 shader APU which is barely better than PS4.

Now an astute AT forum member would have noticed that an HD7850/R7 265 or R7 370 all outperform a GTX750Ti 2GB, but yet in the real world, a Core i3/FX4300 + GTX750Ti would easily outperform the PS4 in most games. Why is that? That's because 1.6-1.75Ghz Jaguar CPU in the current consoles is a huge bottleneck for even an HD7850/R7 265/R7 370 style GPU.

That means the ONLY way for Nintendo to measurably increase GPU performance beyond PS4 is to incorporate a way faster CPU, which will only raise the total cost of the APU and power usage. Unless Nintendo decides to lose $ on hardware, it's a very slim chance for their console to have a much faster GPU than the 1280 shader GCN 1.2/1.3 because going beyond that is only going to worsen the CPU bottleneck.

Realistically speaking, all Nintendo needs is a 896-1152 shader GPU and 8 Jaguar cores clocked at 2.0Ghz and the rest they should focus on other aspects like online gaming, etc.

Maybe they're planning their launch around HBM2? They said next year, which is quite a ways away, so maybe something Greenland based which is supposedly a new architecture, not just a scaling up.

Greenland is supposed to be a lot more efficient; if that's the direction they are headed I doubt power consumption and heat will be their limitation in beating the PS4. Instead cost will be.

 

[Mondozei]

Russian, while you make a plenty of good points, remember that Nintendo has always had a relatively closed ecosystem.
For this reason, cross-platform compability is less of an issue than it is for X1/PS4, since most games will come out on both.

And if Nintendo chooses an x86 arch, which is most likely, it won't be very hard for devs to optimise for higher graphical fidelity on the console in the same way they are doing on PC on decent ports.

I personally think we will see an AMD APU based on 14/16 nm. The ideal would be Zen on 16 nm, with Artic Islands GPU. That'd be amazing, but if the rumors about delays are true then increasingly less probable. Still, even other options would open up for a very decent performance gain. AMD would also need a cash injection, so I don't think Nintendo would struggle that much in order to get good terms on a contract with AMD.

 

[Techhog]

Maybe they're planning their launch around HBM2? They said next year, which is quite a ways away, so maybe something Greenland based which is supposedly a new architecture, not just a scaling up.

Greenland is supposed to be a lot more efficient; if that's the direction they are headed I doubt power consumption and heat will be their limitation in beating the PS4. Instead cost will be.

Nintendo never uses the newest technology. I'll be surprised if they even go for GCN 1.0.

 

[jpiniero]

And if Nintendo chooses an x86 arch, which is most likely, it won't be very hard for devs to optimise for higher graphical fidelity on the console in the same way they are doing on PC on decent ports.

Now that I think about it, it almost has to be ARM, even above PowerPC (which I still think is possible). No way would Zen be possible to do in an handheld any time soon; and the two consoles seem likely to be compatible. They would get backwards compatibility too with the 3DS which is also ARM. And then you wouldn't have to worry about AMD losing the x86 license.

Didn't AMD say they were only expecting about 1B in revenue during the life of the contract? The processor has to be rather dirt cheap too.

 

[ThatBuzzkiller]

Nintendo never uses the newest technology. I'll be surprised if they even go for GCN 1.0.

They will ...

In fact, if anything Nintendo will most likely go with GCN gen 3 since it's an already established GPU micro-architecture ...

What I want to know is how many low level features are they going to expose ?

 

[Shift_]

Zen APU debuts 2016 in NX? One can only dream but Nintendo isn't the type to put beefy hardware into their consoles. Would be a complete reversal of their console approach.

Perhaps since the Wii came out...but historically in the past Nintendo has done well to put good hardware into their consoles.

Gamecube vs Ps2 vs Xbox the Xbox was the clear winner specs wise, but the Gamecube was more powerful than the PS2.

N64 vs PS1, the N64 was much more powerful than the PS1, however, it was severely handicapped by the cartridge format.

SNES vs Genesis...once again the SNES was more powerful than the Genesis

NES v (well to be honest I had an NES, but I do not have any idea what it actually competed with)

 

[HurleyBird]

SNES vs Genesis...once again the SNES was more powerful than the Genesis

NES v (well to be honest I had an NES, but I do not have any idea what it actually competed with)

SNES vs. Genesis... it was more "slower but with better features" (better colour depth, better graphical features, more advanced sound chip, ability to have co-processors on cartridges, but slower CPU)

NES competed with the Master System which was far more powerful specs-wise.

But yeah, I agree that going for a more powerful console than PS4/XB1 would be a good and welcome move.

 

[Techhog]

SNES vs. Genesis... it was more "slower but with better features" (better colour depth, better graphical features, more advanced sound chip, ability to have co-processors on cartridges, but slower CPU)

NES competed with the Master System which was far more powerful specs-wise.

But yeah, I agree that going for a more powerful console than PS4/XB1 would be a good and welcome move.

All of the people who say that they would buy the console if it were a $500+ superconsole would also buy it if it were weaker, or are just lying to themselves. It's pointless.

 

[AtenRa]

Why have more performance than PS4 ??

I would say it is much better to have the same perf at lower cost and smaller console size.

An 8x Puma core + 12x GCN 1.2 GPU cores (768 SPs) + HBM 1 or 2 at 14nm/16nm FF could be more than half the size of the PS4 Chip.
1. More than 60% less Power = smaller less expensive Heat-sink,
2. It would make a very small Console box due to smaller + less expensive motherboard platform.
3. Performance could be higher than XBone and very close to PS4.
Every Console game would be very easily ported and would make a very nice alternative to XBone and PS4.

The cost of the chip may be the same as the PS4 chip at 28nm but the BOM of the console will be lower. Also it will be more appealing due to smaller size, less power and lower heat/noise vs XBone and PS4.

 

[BigDaveX]

NES competed with the Master System which was far more powerful specs-wise.

It also competed with the Atari 7800, which was a bit more powerful on paper (slightly faster CPU, twice the amount of RAM), but used a weird graphics chip that wasn't suited to the direction games were going in. The Master System was indeed a lot more powerful, but then it came out three years after the NES, so it really had no excuse not to be.

 

[StrangerGuy]

Nintendo never uses the newest technology. I'll be surprised if they even go for GCN 1.0.

Of course it has to be some awful custom designed CPU/GPU combo; otherwise they won't be called Nintendo.

I won't be surprised one bit if the 3DS successor has anemic SoCs and sub-720p TN resistive touchscreens selling for $250+ when even ~$150 phones made in some China backyard will have at least quadcore A53s and 1080p IPS screens by then.

 

[Techhog]

Of course it has to be some awful custom designed CPU/GPU combo; otherwise they won't be called Nintendo.

I won't be surprised one bit if the 3DS successor has anemic SoCs and sub-720p TN resistive touchscreens selling for $250+ when even ~$150 phones made in some China backyard will have at least quadcore A53s and 1080p IPS screens by then.

The screen will definitely be sub-720p; probably 480p. They're not going to make that pricing mistake again though, and IPS is possible considering that they used IPS screens (though low-quality ones) for the top screen of the New 3DS XL.

 

[Madpacket]

Based on the response from Takashi I take it that the "dev" kits sent out are just a representation of the performance of what the console will be when the technology is ready sometime next year, not the actual consoles. This is typical of development kits and I believe MS and Sony both used regular PC's with standard hardware that would match the expected performance of the final consoles (PS4, XBOX One).

Mostly likely this means the development kits are some sort of HBM Fury GPU's combined with a high end Intel processor (Or whatever Intel equivalent is expected to perform at Zen processing levels).

I don't know why everyone here expects Nintendo to release more of the same underpowered crap since the Wii / Wii-U. The Wii-U is a commercial failure even though it arguably has the best games of any console right now (too many dude bro's buying up games like Destiny and Call of Duty keep people away from it IMHO) but it clearly can't compete with MS or Sony's offerings on a technical level.

Nintendo is changing things up. They can't do more of the same. The Wii was "lightning in a bottle" as already stated and can't be repeated using the same type of risky approach as the Wii-U has shown. With Iwata dead and plans of massive expansion and streamlining development practices why is it such a surprise that they release a modern console with the latest technology available?

Nintendo wants the same "dude bro's" buying up their consoles so they can play the latest crap from Ubi or Activision. Nintendo has to design a console that's easy to develop and port to in order to win back third party publishers, they can't do it all on their own as the sales have shown.

 

[Techhog]

Based on the response from Takashi I take it that the "dev" kits sent out are just a representation of the performance of what the console will be when the technology is ready sometime next year, not the actual consoles. This is typical of development kits and I believe MS and Sony both used regular PC's with standard hardware that would match the expected performance of the final consoles (PS4, XBOX One).

Mostly likely this means the development kits are some sort of HBM Fury GPU's combined with a high end Intel processor (Or whatever Intel equivalent is expected to perform at Zen processing levels).

I don't know why everyone here expects Nintendo to release more of the same underpowered crap since the Wii / Wii-U. The Wii-U is a commercial failure even though it arguably has the best games of any console right now (too many dude bro's buying up games like Destiny and Call of Duty keep people away from it IMHO) but it clearly can't compete with MS or Sony's offerings on a technical level.

Nintendo is changing things up. They can't do more of the same. The Wii was "lightning in a bottle" as already stated and can't be repeated using the same type of risky approach as the Wii-U has shown. With Iwata dead and plans of massive expansion and streamlining development practices why is it such a surprise that they release a modern console with the latest technology available?

Nintendo wants the same "dude bro's" buying up their consoles so they can play the latest crap from Ubi or Activision. Nintendo has to design a console that's easy to develop and port to in order to win back third party publishers, they can't do it all on their own as the sales have shown.

Better specs won't sell consoles for them. Their choices are to make a very cheap console, or to go third-party. Anything else is guaranteed to fail and cost them billions.

 

[poofyhairguy]

why is it such a surprise that they release a modern console with the latest technology available?

Because that is a huge risk for a company that has tried to minimize risks since the Nintendo Wii. If Nintendo puts out a $200 ARM console and it's not a huge success then they probably still made money per unit and they are set to move to a Sega role with easy ports of all that console's games to mobile devices. That path has a clear future. If Nintendo tries to push a top of the line console- aka the kind where they lost a little money per unit the first year or so- if it flops they are FORCED to move to a Sega role without billions left in the bank. That path is unclear.

I mean, the Japanese market is pretty much gone so Nintendo can't rely on that. To compete with the PS4 or Xbone straight on means Westernizing everything to appeal to the largest market that is left. To beat either company at their own game would require some cash spent on the hardware, online play and marketing. You pretty much need Nintendo to be bought by someone like Apple for them to have those kinds of resources. Moving forward with a top tier console is a huge risk without that sort of backing because they probably will have to cut a leg off of that stool (no decent marketing, or decent online play for example).

Plus there is the fact that in order for a console to succeed Nintendo needs help from the third party developers that it has pissed off since the 1980's. They can barely get out simpler Wii U games on schedule now, there is no guarantee that their first party games can carry a more powerful console when each game has an even LARGER budget and resource commitment. And they need more than ports to make their console a real value proposition, they need exclusives made either by them or a third party. That is asking a lot out of 2015's Nintendo.

Honestly I will almost be sad if Nintendo does make a PS4 killer. It will be the company's Dreamcast almost certainly. Give us the $200 ARM box Nintendo and stay alive another decade.

 

[Headfoot]

Nintendo being Nintendo we can count on whatever the end chip is being loaded with fixed function silicon for a variety of purposes. I would guess fixed function accelerated audio at the minimum, custom interconnects, custom cartridge/disc tech again (the mass market is close but not ready for a digital download only mainstream console). Maybe even copies of all fixed function tech from their mobile offering so they can be directly compatible with Mobile on home Console.

I would also guess that there will be focus again on a custom input or output (wiimote / wiiu tablet / 3ds 3d). There is a chance it could be VR related, or maybe another take at 3d on mobile. I dont think it will be VR however, because the computation demands are probably still too high for mainstream consoles.

My bet is on a semicustom ARM CPU (a72 derived or even more custom than that). ARM CPU lets them do cool cross compatibility stuff with their Mobile offering, and Nintendo has done a Mobile+Console pairing of some variety for every console since NES. Count on seeing it again. If not ARM, Zen, but I don't think that's as likely.

I think there is a possibility however of a customized and updated Jaguar++ with lots of fixed function silicon with PS4 equivalent GPU performance. If you asked me 3 years ago if Nintendo would ever break down and put their mobile franchises on Android/iPhone as apps I would have said no. But lo and behold they are releasing an "endorsed third party" Pokemon app. Times are a changing.

I think the most concrete part is probably the GPU being some sort of GCN derivative

 

[Techhog]

Nintendo being Nintendo we can count on whatever the end chip is being loaded with fixed function silicon for a variety of purposes. I would guess fixed function accelerated audio at the minimum, custom interconnects, custom cartridge/disc tech again (the mass market is close but not ready for a digital download only mainstream console). Maybe even copies of all fixed function tech from their mobile offering so they can be directly compatible with Mobile on home Console.

I would also guess that there will be focus again on a custom input or output (wiimote / wiiu tablet / 3ds 3d). There is a chance it could be VR related, or maybe another take at 3d on mobile. I dont think it will be VR however, because the computation demands are probably still too high for mainstream consoles.

My bet is on a semicustom ARM CPU (a72 derived or even more custom than that). ARM CPU lets them do cool cross compatibility stuff with their Mobile offering, and Nintendo has done a Mobile+Console pairing of some variety for every console since NES. Count on seeing it again. If not ARM, Zen, but I don't think that's as likely.

I think there is a possibility however of a customized and updated Jaguar++ with lots of fixed function silicon with PS4 equivalent GPU performance. If you asked me 3 years ago if Nintendo would ever break down and put their mobile franchises on Android/iPhone as apps I would have said no. But lo and behold they are releasing an "endorsed third party" Pokemon app. Times are a changing.

I think the most concrete part is probably the GPU being some sort of GCN derivative

Sorry about nitpicking, but The Pokemon Company (a subsidiary jointly owned by Nintendo and Game Freak which is pretty much self-run) is releasing that app, not Nintendo. Also, it's not like it's a full-fledged game or even a hint that there might be one someday.

 

[sxr7171]

So many people keep saying that Nintendo needs to release a console way more powerful than XB1/PS4 but I disagree for the following reasons:

1) Let's say just throw it out there and assume NX is 3-4X faster than PS4/XB1. Would that matter? No. The reason it wouldn't matter is that by the time the NX releases in 2016, XB1/PS4 will have an install base of at least 50 million units. The NX will be a new kid on the block. Why would 3rd party developers spend extra effort to take advantage of much more powerful Nintendo hardware when there will be so few NX owners in the first 2-3 years of the console's life? After all, even hardware 3-4X more powerful isn't just going to sell the console when "most of your friends have XB1/PS4 or both and console games like playing with their friends." It also means the NX will have to contend with by then a huge library of games to choose from among XB1/PS4 consoles. More powerful hardware also has to go hand-in-hand with Nintendo actually getting more 3rd party support, which in itself is doubtful given the track record of Nintendo's 3rd party support since N64. If they can't secure a lot of 3rd parties, what's the point of all that extra horsepower? This is because Nintendo itself isn't the type of company that spends hundred million dollars making the most gorgeous game and honestly I doubt their team even has the technical ability to do so. We wouldn't suddenly see the next Zelda, Mario, F-Zero game blow the doors off Uncharted 4, Rise of the Tomb Raider, Gran Turismo 7 because Nintendo isn't a firm that knows how to make the best looking 1st party games either. In simplest terms, given how underpowered the Wii U is, even a GPU at XB1 level is more than enough for Nintendo's game designers to make much better looking games. They don't need the power of the Nano or 980Ti in there.

2) Adding a GPU far more powerful than XB1/PS4 would mean a more expensive console since Nintendo is reluctant to lose $ on hardware. By end of 2016, it's possible there will be even more price cuts on PS4/XB1. Nintendo cannot afford to launch a $450-500 console with 3-4X the power of PS4 when by 2016 it'll be easy to pick up XB1/PS4 for $299. This is another reason going with a powerful Nintendo GPU is too risky.

Another way to look at it is this -- if the next Nintendo console is much more powerful than the existing consoles, by the time we would start to see some games taking advantage of that hardware, it'll be 2-3 years and by 2018-2019 we could be looking at XB2/PS5 which could be 6-8X more powerful than XB1/PS4 and again trounce the 2016 Nintendo console.

For that reason, Nintendo needs to focus on offering something unique or focus on low price. The strategy of a powerful Nintendo console would have been far more successful if they launched it in 2013 but not mid-cycle in the generation.

I would honestly be surprised if the next Nintendo console is actually more powerful than the PS4.

I agree and this is why I don't understand why they need a new console. If anything they could have made some proper games for the WiiU. So some of us bought this thing and the only thing of note we got was that trio of: 1. A new 3D Mario platformer 2. SSB 3. MK8.

The only one of those I really liked and cared about was Super Mario 3D. It was great but not worth $299 + $60. There is an implicit promise when you buy a Nintendo system that you will get some Nintendo franchises. At least ONE new Zelda game and probably a new Metroid game.

Suddenly they abandon the whole platform without even one new Zelda game? I feel like they robbed me. Why the need for a new system all of a sudden? To milk us again? I'm not buying from Nintendo ever again. They threw in the towel on a perfectly good system.

This is the beginning of the end of Nintendo and I'm sure of it. It reminds me of what happened to Sega when they disappointed Saturn fans. Even though they followed it up the glorious Dreamcast they still lost.

I don't except anything from Nintendo to come even close to how good the Dreamcast was for its time. It's going to be something rehashed and something gimmicky. They seem to try different gimmicks every generation like throwing darts at the board blindly and hoping one hits the bullseye.

They've burned 3rd parties over the years but this time they burned their fanbase. This next system of theirs will be their last. They might as well just pick up a PS4 dev kit right now and start making that new Zelda on it. As for the next rehashed MK and SSB they should run fine on a Roku or an Apple TV. I suppose to keep this on topic their next console will likely be ARM based and if they want to go with x86 they might as well start developing for an existing console.

 

[Madpacket]

Better specs won't sell consoles for them. Their choices are to make a very cheap console, or to go third-party. Anything else is guaranteed to fail and cost them billions.

Of course better specs will help them sell consoles. I bought a PS4 over an Xbox One for this reason (among other reasons, but this was a major factor). I don't think I'm alone here.

Nintendo still has a very large stockpile of money to pull from. This quote is a little old but still relevant.

"Buried in reams of financial data is the revelation that Nintendo have 812.8 billion Yen (£6.7/$10.5 billion) in the bank - enough for it to take a 20 billion Yen loss (£163/$257 million) every year until 2052.Feb 13, 2012"

They made a profit last quarter, they're not bleeding money. They're not going anywhere, they've been around for 150 years!

Nintendo's challenge is to win back third party developers to make games for their console(s). That's really it. It's a big challenge sure but If they can do that on a game console that outperforms the PS4 and do it at competitive pricing they will sell like crazy.

Remember they have the best first party games and developers in the business. Imagine what they could do with a game console even twice as powerful as a PS4 (this should be easily achieved with a HBM/APU at 14/16nm).

They can't really go ARM (unless mobile) or PPC, they have to go X86 to be on even playing field with Sony and Microsoft. Backward compatibility will probably be thrown out unless they can shoehorn a PPC chip into their design but this would add to cost and is really unnecessary given PPC emulation.

Sure someone could buy them but they would probably cost as much as Disney so good luck with that.