basically what I want from Nintendo next-gen is what they have already been hinting at lately. a few statements from Nintendo include, "next console and handheld will be more unified like brothers" and "the transition from GC to Wii was the easiest we ever had and hope to replicate that next-gen by absorbing the Wii U architecture".

basically what that means to me is that Nintendo's next handheld and console will share much of the same hardware which will allow them to play the same games as one another.

I believe the handheld version will be somewhere between Vita & Wii U in power like 3DS is between PSP & Wii.

the console version will essentially be a super Wii U, similar to how Wii was basically a super GameCube. the extra boost in power over the handheld version will allow games to run at a higher resolution/frame rate.

these devices will launch sometime between holiday 2016 & holiday 2017, possibly with staggered launches. initial price $199 each. they will share the same virtual console/eshop and support cross-buy/cross-save so if u own a digital game on one u will own it on the other as well. also they will have a unified online system so someone with the handheld version can play online against someone with the console version. they will have some sort of connectivity as well so a handheld can act as a controller for the console like the Wii U gamepad.

the most important part of this idea is that it allows Nintendo to focus all there attention on one single platform which will hopefully allow them to supply a steady schedule of games, something they have struggled to do while supporting two seperate device. this also allows Nintendo's home console to regain strong 3rd party support since there handhelds always have strong Japanese 3rd party support and a ton of western games based on kids shows/movies.

one other thing I would like to see is a much improved Virtual Console along with a subscription service. $50/year gives u unlimited access to the virtual console. this gives consumers access to hundreds of games for cheap and since the average person probably doesnt spend $50 on it per year, Nintendo would be making more money.

since Nintendo can't really afford to go all out against MS/Sony for specs/multimedia features and doesn't get most of the AAA western titles, they need to find a way to differentiate themselves further and be more self-reliant. this scenario is the best way for them to do that and hardcore Japanese games+child friendly western games can easily fill in the gaps.

I don't know why, but I would be very satisfied if it's up to date hardware with new version of gamepad.

I do really enjoy the gamepad, but it should be only used as auxiliar screen(map, quick inventory), get rid of the offscreen gameplay.

In my mind if the wii u was almost as powerful as the Xone, it would sell a lot more than it actually does.

I would try to go the Sony/MS-route with the next console and just try to offer a system without too much gimmicks. It may be that this space is already too crowded with Sony and MS being there, but I don't think that a gimmick-focused console will work either (unless the gimmick is really good).
And Nintendo could launch their next system 2-3 years before Sony/MS follow (if they plan to release next-gen systems at all), which would give Nintendo 2-3 years of having the most graphically capable console. In addition to offering the typical Nintendo franchises the console should be capable of playing multiplatform-games at 1080p60, even if the release on the XOne only offers 720p30. I think that Nintendo should also follow suit and switch to a x86-platform. This would make the development for their new console easier. Nintendos own franchises are enough to differentiate them from the competition.

And I always liked the codenames for their consoles, and despite my proposed console only being an evolution of the current ones I would call the Wii U successor the Nintendo Revolution. Sets the name apart from the Wii label, while still being connected to it, and could be marketed quite easily.

Here are the specs (and some other stuff) I would give the Nintendo Revolution:
- release: Holiday 2017
- GPU: ~3x as powerful in all areas as the PS4 counterpart
- CPU: ~2x as powerful as the XOne/PS4 counterpart, but with only 4 cores (so 4x as powerful per core)
- if this CPU is strong enough to emulate the WiiU-CPU I would add other minor stuff that enhances WiiU emulation (e.g. the 32MB eDRAM from the WiiU as an L3-cache)
- RAM: 16GB stacked RAM @ ~600 GB/s  (with 4GB reserved for the OS and background applications)
- standard gamepad (but compitability to Wii U Gamepad and other similar devices)
- integrated 256GB m.2-SSD @ ~1000 MB/s sequential
- USB 3.1 support and the option to connect an external HDD (I actually liked that decision with the current WiiU)
- HDMI 2.0 support (most games would still render in 1080p60, but higher rendering resolutions would be possible)
- if ~64 GB SD-cards are really affordable at the end of 2017 in addition to digital distribution all games should be delivered via SD-cards (but the cost of a 64 GB SD-card needs to be <10$ to allow this); if not possible a BD-drive capable of reading discs with a size of up to 128 GB should be included

Price: 399$ (the PS4 will probably be around 249-299$ when this console launches); if Nintendo can exclude a BD-drive this might even be profitable on itself

So and now I'm going to contradict myself a little bit: I really hope that VR kicks off. And if it does I would like to have an additional Nintendo Revolution SKU (called the Nintendo Revolution VR/3D Edition), that performs better with the VR-"gimmick" and offers "1440"p72 @ 90° FOV, 2x1080p60 on 3D televisions and 1x1080p72 when used on 72/144 Hz displays. I think that it could be enough to only render the 1920x1080 field in the middle of the VR-display in full resolution. For the outer parts the rendering resolution could be dropped by half in each dimension, reducing the required additional computational power.

Changes to the standard SKU would be:
- GPU with 1.5x as many units, overclocked by 20%
- CPU overclocked by 20% to offer stable 72 FPS
- improved cooling system; system would still be slightly louder than the standard SKU when running in VR/3D mode
- includes "entry-level" VR-headset (maybe in cooperation with Oculus?) with an 1440p72 display and 90° horizontal FOV optics, that isn't limited to usage on the Nintendo Revolution but can also be used on PCs and other devices; other VR-headsets are also supported but anything outside the assumed 90° FOV just remains black and only 72 frames/second are delivered to the headset even if it would support more (except when game developers add support for other resolutions, FOVs and/or refreshrates themselves)
- 16 GB stacked RAM @ ~1200 GB/s (different setup with 8 packages instead of the 4 used in standard SKU)

Price: 599$ (maybe an additional SKU without an included VR-headset for 469$ could be offered)

But as I said, I would only suggest to do this if VR really kicks off. No need for Nintendo to push it. They should just be ready if VR starts delivering (they probably will have to decide at around the end of 2016).

A lot of people asking for tons of computational power are forgetting about cost reduction issues on newer process nodes ...

It'll be a feat if the WII U's successor matches the PS4 in terms of capability seeing as how Nintendo is prone towards aggressively pricing their consoles in the sub $350 bracket and is obsessed with smaller form factors along with lower power consumption.

First, in my dream console i would be able to play handhelds games on the home console.

my dream console is a revolution in console making, and depends on technology to be feasible. they will have to wait more (launch arround 2019/2020).
meanwhile, start talking with ibm(to use the machine) or someone who is managing to use graphene insted of silicon on microchips, and make the console with one of these new technology(they say that this 2 options would be hundred times faster and cheaper). would be more powerfull, cheaper, smaller than ps5(if ps5 dont use graphene also).
hire programers to port the new engines to this new arquitecture meanwhile not launch.

A Gtx 970 is only slightly more powerful than a ps4? Hell naw!

I don't understand the hype behind graphene ...

Other than the fact that it has an extremely good electron mobility which is ideal for higher frequencys but it's not good enough to represent digital signals due to it's low band gap. Bilayer graphene isn't that promising either due to it's issues with low on-off current ratios compared to bulk silicon despite solving the band gap issue. We're at least 10 ten years away from realizing commercial graphene based electronics ... 

Either that, or they are tech savy enough to know that with current technologies you can't go much further than the PS4 without consoming too much power and thus producing too much heat. A Console has a max TDP of something around 150W without creating too much cooling problems (RROD anyone?)

To answer the OPs Question: It would be too soon to decide right now, but I'll give it a try anyway

I would continue with the Gamepad idea, but with a more advanced Pad. 720p or 1080p capacitive Screen (possibly with multitouch) and better batteries would be the Main changes here though, besides allowing to connect one per Player (Max 5). Using a "normal" game controller in Games not needing the gamepads Features should however still be possible. And watching different shows on every screen with Netflix and other Videostreaming sites should also be possible.

Now to the Main Event, the hardware of the console itself. First, the main processor:

Nintendo had a long run since the N64 with IBM POWER Processors and ATi (now AMD) Graphics Chips, so I think this cooperation should be continued. Modern IBM processors however are designed for high power servers, which makes them pretty useless in a Console unless heavy modifications are applied. A modern POWER 8 would need to be severly stripped down in many functions extremly useful in servers but virtually, if not absolutely, useless in a console. But I'm confident taht Nintendo could make something like a 6-core @ 3Ghz out of it (8-core could be pushing it since the POWER 8 cores are much more powerful and thus consuming also much more than the Low Power AMD Jaguar cores used in the PS4/XBox ONE). Should the modern POWER Architecture prove too troublesome, I would suggest adapting the upcoming AMD Zen processors (completly new architecture from AMD for Desktop/mobile processors announced for 2016)

It's still too soon to argue about the graphics, since AMD ist still using pretty much the exact same architecture as they are using in the PS4/Xbox ONE Consoles. Which means if they would create a console right now it couldn't be any stronger then a PS4 in the graphics department no matter what. Guess I'll have to wait until the Pirate Islands (codename of the Radeon 3xx Chips) comes out.

As for the Data storage, I would stick to the Flash Memory. Hard Discs are on the way out and are beginning to get irrelevant in sizes below 1TB because the price tags of SSD (who consists of Flash Memory) are plumetting and will continue to do so while HDD are getting more costly because of the diminishing number of producers and ever more complex production (models above 4TB for instance need to be filled with helium to reduce friction). One can get a 250GB SSD for less than 100€/$ nowadays, and I'm pretty sure in 2 years the price tag for such a drive will be less than 50 and terabyte SSDs will be approaching the 100€/$ mark very fast. Plus, SSDs are way faster than HDDs ever could posibly get.

More revolutionary on my part would be that I would switch back to cartridges, albeit much smaller ones (think somewhat enlarged GBA cartridges). The reasoning behind this is that Blu-Ray has too many limitations. It is already beginning to become too small, on PC some games already take +50GB of hard disc space, more than twice a single layer Blu Ray disc can handle. One could increase its storage by adding some more layers but then the transfer speed of just around 40MB/s will begin to cause serious problems, as loading times would get endlessy long. A cartridge however is scalable in storage size and has very fast transfer speeds (in the GB/s region), shattering the limitations of optical discs. They do have a drawback however, and that is the fact that they are more expensive than any optical disc and more difficult to produce, so games might get a bit more expensive due to this.

As for connections, I would add 6 USB 3.1 up front to load the gamepad's batteries and connection of a USB 3.1 compatible stick (3.1 is physically different to its predecessors and thus not compatible without an adapter), and 2 3.0 in the back for external devices like hard drives. And a connection for an internet cable, Wi-Fi can be VERY troublesome sometimes.

And finally, install dust filters!

Phew, turned out to become a much longer post than I expected for my very first post here. So, hi everybody ;)

Almost everyone's console is gonna flop and be ignored by 3rd parties 12moths after it launches. why do you guys have such low expectation? lol

A GTX 970 is very expensive to manufacture coming in at about 400mm^2 die. The PS4's whole entire APU which also consists of it's CPU has a die area of 330mm^2. What's more is that not a whole lot of engineers are expecting cost reduction anymore so it's very unlikely that a GTX 970 will go for less than $260 in the near future ...