What the Wii U hardware should have been?

[Proelite]

This is a discussion that I had with a friend of mine (huge Nintendo fan) who agreed that hardware is the biggest problem for Nintendo for the foreseeable future.

Both us agreed that the Wii U could have been a more viable product had it not included the gamepad and hardware BC. The reasoning being that gamepad failed to light the world on fire like the Wii motion controls did, and the BC forced Nintendo to reuse CPU architecture that hindered easy of porting and possibly performance at the cost of additional investment.

Anyways, in our opinion the Wii U should have been

4 core Piledriver at 1.6 - 2.0 ghz
256-512 GCN cores at 400-600 MHz
32mb of edram + sram
Audio DSP

4 gigabytes of DDR3-1600 on 128-bit bus

Split motion controllers, 2 "enhanced" nun chucks with buttons.

TDP 50 Watts, 100 watts PSU
Cost:$299 base

Benefits:
Plays current gen ports at 1080p and/or 60fps, thereby becoming the multiplatform of choice for at least a year.
Capable of running next gen ports at 720 and/or 30fps with ease
Nintendo exclusives that rivals Xbox One and PS4 games in graphical fidelity.

Negatives:
No BC
No outstanding difference compared to Xbox One and PS4.

Would the increase in hardware power make a positive difference in the long run?

 

[liolio]

I think the WiiU hardware should have been good

------------------------>[]

 

[Proelite]

I think the WiiU hardware should have been good

------------------------>[]

Liolio, I made this thread just for your opinion.

You always write good entries for hypothetical what-if scenarios.

 

[BRiT]

Would the increase hardware power make a positive difference in the long run?

No. I think Nintendo's real failings are the lack of proper support for third party developers as well as their own failings on the user experience. No one but the extreme faithful are going to buy a hardware product just for the Nintendo franchises. The entire Nintendo online experience is extremely anemic even when compares to PSN at the time of PS3's launch. Nintendo had years to refine the experience and failed utterly to do anything positive during that time.

 

[Proelite]

No. I think Nintendo's real failings are the lack of proper support for third party developers as well as their own failings on the user experience. No one but the extreme faithful are going to buy a hardware product just for the Nintendo franchises. The entire Nintendo online experience is extremely anemic even when compares to PSN at the time of PS3's launch. Nintendo had years to refine the experience and failed utterly to do anything positive during that time.

Won't the hardware in the OP fix that? It's basically a ultra high-end laptop APU / mid-end version of Xbox One. Porting should theoretically be extremely easy.

 

[BRiT]

Won't the hardware in the OP fix that? It's basically a high-end laptop / mid-end version of Xbox One.

No, not unless they basically use something drastically different for developers than what they offer up on the Wii U. Nintendo has software issues. Their developer tools and documentation is extremely anemic. Their entire OS is anemic. Their online network infrastructure is anemic.

 

[Proelite]

No, not unless they basically use something drastically different for developers than what they offer up on the Wii U. Nintendo has software issues. Their developer tools and documentation is extremely anemic. Their entire OS is anemic. Their online network infrastructure is anemic.

Ah...

That's the first I've really heard about Nintendo having poor developer tools. That would explain a lot.

 

[Grall]

No. I think Nintendo's real failings are the lack of proper support for third party developers as well as their own failings on the user experience.

Part of the reason that Nintendo's failed with 3rd parties is that they designed a last-gen console seven fucking years after last-gen started, and 3rd parties aren't interested in continuing to design last-gen games when they're busy re-gearing for making next gen games.

Nintendo's biggest friggin fail is that they didn't make a console that at least could PRETEND to be within spitting distance of next-gen consoles, even if it only reached 60% of PS4 or xbone. Now it isn't even 20%, so it won't be able to run the games that are in the pipe for next-gen consoles, even if significantly cut-down.

A wuu that was reasonably similar performance to 4bone and say, 4GB RAM instead of 8 could have run the same game with reduced textures and lower rez. Wuu...? No fecking way. It's not even DX11 for chrissakes. No SIMD in the CPU cores, and so on.

Nintendo had years to refine the experience and failed utterly to do anything positive during that time.

Yeah that's what really gets my goat, how Nintendo could spend so damn much time coming up with a console filled with nothing but horrible hardware, the world's ugliest, slowest shittiest menu system and absolutely NO GAMES WHATSOEVER for it. It's like they totally forgot everything they learned over the course of 30 years about how to launch a console system.

 

[Rangers]

Would the increase in hardware power make a positive difference in the long run?

probably not, because it's still in the end quite below xb1/ps4.

well, i guess depends what you mean by made a difference. it would have put up much more of a fight imo, but the ultimate outcome the same. By your definition "in the long run", then no.

also piledriver cores are hot and big and dont jive with any modest design. better 4 jaguars (clock them at 2ghz even maybe)

 

[swaaye]

I don't think they need supar 3d powaa parity. But if not that then they need some other draw. Motion controls are not thrilling like in 2005 and the tablet is uninteresting. So what's going to get 3rd parties or potential customers to care? Nintendo is out of touch. They've been that way forever though. For one their culture is bizarre and not friendly to developers and never has been.

 

[warb]

The tablet controller was not the best idea. WiiU with improved waggle and a Kinect/Eye like camera would've killed it.

 

[kalelovil]

I'd argue that instead of changing the specifications, releasing it two years earlier would have helped Nintendo more.
I don't see an obvious internal hardware-related reason it couldn't have been released in 2010.

 

[Esrever]

If nintendo is going to be this far behind on hardware. They need to cut their console life cycles smaller. If the Wii U announces in 2010 and launches in 2011, they could have had the same performing console as they have now with less competition from the next gen, then they can replace the Wii U in 2016 with hardware exceeding the ps4 and be great. They can sell the extras as add on instead of the full console to get better pricing.

This nintendo could sell the console as an alternative to the other ones instead of their own unique thing which may or may not catch on. All this they should do with BC included so it will be a smooth transition.

 

[Tchock]

AMD doesn't license Piledriver out (nor is it easy to port out to other non GloFlo fabs).

If they released before '13, they'd probably had to go for Bobcat, which was only available for 2-core clusters at 40nm. Not that great a choice, either.

 

[dagamer]

Broad strokes:
Quad core CPU, GCN GPU, 4GB DDR3, 1GB GDDR5, wired Ethernet, 250GB hard drive. $299.

No tablet. WiiMote HD with built-in Wii Motion Plus and force feedback nunchuck. 9-axis gyro.

Oh, and released in 2011.

 

[V3]

I don't think there is anything wrong with the specs of Wii U hardware. Nintendo can't be another Xbone or PS4.

What Nintendo should have done is integrate the Wii U into the WiiUPad and turn it into a stand alone gaming tablet that can output to the TV if needed. Their mistake was to chase HD graphics by making Wii U stand alone unit. The living room is a dead end for them IMO. It's a dead end for Sony and MS too I reckon.

Tablet gaming is a growing market and something that is worth chasing for Nintendo. A gaming tablet with 8-10" screen and decent graphics capability, battery life and ergonomics would have been a better product compare to Wii U. They can even add a controller shell shapes like the WiiUPad that you can put the tablet into if buttons are important to them.

 

[warb]

Yes, it wasn't never a bright idea to split attention between tablet and TV, they should have picked one or the other.

 

[liolio]

Liolio, I made this thread just for your opinion.

You always write good entries for hypothetical what-if scenarios.

I honored, sir.

I think that we discussed that topic a lot already in other, inappropriate, threads and the more I think the more I believe that Nintendo did the best it could with the hardware based on the business decisions that were made about specifications and timelines.

I wrote a long post before this one and discard it, too long, not right on the money if not... confused...

I though of what for example Chinese integrator could have done for a system that would have cheap @250$ it would have integrated a "miniWiiUmote" with it own resource that could have shipped for 100$. I won't get into the details but as I was writing I realized that whereas having a controller with its own resource would have allowed for more interesting scheme on top of the one Nintendo went with but it would have gotten them to compete against their 3ds.
So not an option.

I was let with a system ARM/PowerVr based, without BC, made on TSMC 28nm process, that would have been in the ballpark of the PS360, most likely superior and could have shipped for cheap, like 150$, with wiimote+. Cheap enough to leverage impulse buy but the lack of BC could have bothering for one in need for a replacement for their Wii, and a bit late and offering notion new to either casual or core gamers.

So I'm left with 2 massive issues, the WiiU is too late and the 3DS set restriction to what Nintendo can even if they were to change their business practices.
It is even more bothering when you think that looking at declining Wii sales, the 3DS and and the Wii2 should have launched hand in hand, when the Wii still had some traction.
Launching meant using a 40nm process (ultimately Nintendo did... in 2012) which made it tougher to offer a better PS360 design to ship at a really low price, still doable.
BC could have been handled like in Durango with an HDMI in port, later on through emulation.

Then you have the issue of the network infrastructure, Nintendo should have contract the proper people to build and use something that can scale depending on their success and user base. Imo they should have gone with either MSFT. Google or Amazon cloud resources.
They should have contracted people to work on the OS. Again it is a complete change in business practices. MSFT has a sane OS which takes 32MB, Sony got down the PS3 OS footprint to 50MB, imo 1GB for the OS is a terrible joke... what a waste...

In any case Nintendo did not have an easy success ahead, they also face issue on the software side, their main IPs are always late, etc.
Even if you launch 2 years ahead of the competition, something really affordable that can trigger impulse buy, you need some software to be ready.

Overall the company is not managed properly, it is not new, the WiiU takes a lot of criticism but imo it is partly here because of 3DS. Nintendo needed something new to push the system, the second screen was an idea and they don't have the tech to match either eye toy or Kinect.
How the 3ds turned out show that one year ahead of the WiiU launch Nintendo still did not have a decided about what would be the selling point of the WiiU. In that context the WiiUmote looks like an after thought.

I think the WiiU is just a symptom of Nintendo inner "troubles". With their business paln the WiiU was may be the best one could do.
They wanted BC, I think IBM knew that much. IBM may have said something like this "we are interesting in doing something for you but it has to use our technology" that reads you have to use our foundry capacity. Even if IBM made them a nice offer that tiny CPU may cost more then it looks.
It has other implication, a bigger CPU is not an option, and an APU out of the picture altogether.
Now what to do with that CPU, I guess Nintendo did not wanted to face the limitation of the EDRAM in the 360 and decide to integrate the the EDRAM to the GPU. Another "boutique" process and quite likely a chip bigger than its size implies. But what could they do?
A 128 bit bus+DDR3 may have performed worse or not significantly better (AMD APU are doing OK at 720P but we speak of big hot and most likely to expansive chips).
I guess BC also had an impact here on the GPU design if not significant in performance but quite an overhead in R&D.

So what the WiiU should have been? It should have been one product in the portfolio of products of a company that know where they are going, that adapts to a changing environment, that has good business practices. A company that thinks that games are still toys, that think they should be affordable but that does what it takes to live up to its own standards.

 

[-Sweeper_]

In my opinion Wii U should have been:
Custom 40nm AMD APU
4x Bobcat cores @ 1.6GHz
400-480 ''Evergreen/HD5xxx era'' VLIW5 stream processors or 384 VLIW4 ''NI/HD6xxx era'' stream processors @ 700-800MHz
4GB DDR3 @ 1600MHz
32Mb eDRAM
250GB HDD
U$249 (and a stripped down U$199 SKU before PS4/X1 launch).

That would make a small, power efficient, cheap yet very capable system.

 

[pc999]

Add a extra CPU core just for the sake of it and "innovative" wii remotes. Love the low power consumption.

I am fine with the console just needs to be cheaper