Nintendo GOing Forward.

[liolio]

And why are you assuming that AMD isn't capable of delivering an ultra-mobile SoC?

I'm assuming nothing, AMD X86 lines are not compliant for mobile usage. In the future? May be, may be not.
Wrt to ARM CPU is is unclear what AMD plans are but my understanding of their strategy is that they aim at low end servers more than mobile parts they seem unwilling to compete with Qualcomm, Samsung, etc. or even Nvidia. I guess the point is that overall their strategy going forward is quite nebulous be it ARM or X86, actually even their GPU roadmap does not give much visibility.

Why focus on AMD? That is pretty much binding oneself to a manufacturer as Intel is likely to remain too expensive. If X86 is out there are other valid options.

Anyway I don't think their main issue is with hardware, their hardware decisions may have been subpart but it is not the harder part to fix.

 

[Grall]

Why on earth would Nintendo go with x86 (well, x64 these days) for their next console? That's...nuts.

The only absolute reason you have to choose intel architecture is if you have a library of old software you need hardware compatibility for. Since nintendo hasn't used any intel CPUs since forever (well, maybe in one of their old arcade systems, I dunno), that's not an issue. Intel still does not offer superior price/performance, nor do they beat competitors on power consumption, after how many years of trying?

Nintendo has basically no prior experience with x86, but plenty with ARM and PPC, some with MIPS as well. I think the choice is quite clear, really.

 

[-tkf-]

Why on earth would Nintendo go with x86 (well, x64 these days) for their next console? That's...nuts.

Cheap and power full , if from AMD?

 

[McHuj]

The only reason I could see Nintendo going with x86 is for easier porting capability between the next console and X1/PS4. But for that to work, I would expect them to make drastic changes in how they approach and communicate with third party developers. It would require a culture change at Nintendo so I doubt that will happen.

What I think Nintendo should do is go with an existing 28nm SOC and not develop one of their own, something like a chip an existing Snapdragon line. It's probably already more powerful than what they would build, they're not on the hook for manufacturing as they will basically be buying a commodity part. You already have built-in Wifi (even LTE), camera, and all the peripheral connectivity they would ever need into the SOC. I think the net BOM would be lower than with a custom built component.

 

[I.S.T.]

Why on earth would Nintendo go with x86 (well, x64 these days) for their next console? That's...nuts.

The only absolute reason you have to choose intel architecture is if you have a library of old software you need hardware compatibility for. Since nintendo hasn't used any intel CPUs since forever (well, maybe in one of their old arcade systems, I dunno), that's not an issue. Intel still does not offer superior price/performance, nor do they beat competitors on power consumption, after how many years of trying?

Nintendo has basically no prior experience with x86, but plenty with ARM and PPC, some with MIPS as well. I think the choice is quite clear, really.

So, Sony going with x86 was a bad decision?

 

[Shifty Geezer]

Why on earth would Nintendo go with x86 (well, x64 these days) for their next console? That's...nuts.

Cross-platform. Unless Nintendo's strategy is to be the second console, and not a direct competitor for PS4/XB1, being easy for developers is very important. x86 and PC GPU affords immediate, cheap, easy targeting by developers, which is key for Nintendo to ever be more than a Nintendo-only player.

Alternatively, go with PPC or MIPS and awkward in-house tools and costly ports to a smaller userbase scaring away already apprehensive publishers, and just run some super cheap Nintendo-only box with 3 games a year from 1st parties struggling to come to terms with their own hardware.

 

[Deleted member 13524]

Why on earth would Nintendo go with x86 (well, x64 these days) for their next console? That's...nuts.

The only absolute reason you have to choose intel architecture is if you have a library of old software you need hardware compatibility for. Since nintendo hasn't used any intel CPUs since forever (well, maybe in one of their old arcade systems, I dunno), that's not an issue. Intel still does not offer superior price/performance, nor do they beat competitors on power consumption, after how many years of trying?

Nintendo has basically no prior experience with x86, but plenty with ARM and PPC, some with MIPS as well. I think the choice is quite clear, really.

Hey, look at a quote from alternate-Grall, back in April 2013:

Why on earth would Sony go with x86 (well, x64 these days) for their next console? That's...nuts.

The only absolute reason you have to choose intel architecture is if you have a library of old software you need hardware compatibility for. Since nintendo hasn't used any intel CPUs since forever (well, maybe in one of their old arcade systems, I dunno), that's not an issue. Intel still does not offer superior price/performance, nor do they beat competitors on power consumption, after how many years of trying?

Sony has basically no prior experience with x86, but plenty with ARM and PPC, some with MIPS as well. I think the choice is quite clear, really.

alternate-Grall was wrong a year ago, so you might be wrong too.




x86-64 and GCN from AMD would be for instruction parity with the other consoles and PCs.
A mobile console using a SoC from AMD would be within Nintendo's announced plans of using the same hardware platform between mobile and home console.

 

[Fourth Storm]

I'm assuming nothing, AMD X86 lines are not compliant for mobile usage. In the future? May be, may be not.
Wrt to ARM CPU is is unclear what AMD plans are but my understanding of their strategy is that they aim at low end servers more than mobile parts they seem unwilling to compete with Qualcomm, Samsung, etc. or even Nvidia. I guess the point is that overall their strategy going forward is quite nebulous be it ARM or X86, actually even their GPU roadmap does not give much visibility.

Why focus on AMD? That is pretty much binding oneself to a manufacturer as Intel is likely to remain too expensive. If X86 is out there are other valid options.

Anyway I don't think their main issue is with hardware, their hardware decisions may have been subpart but it is not the harder part to fix.

I agree they have much bigger problems than hardware. I believe there is room in the market for a low/medium end console, but Nintendo just botched the execution with Wii U.
ARM64 would be cool in their next hardware refresh, but from what I've read, you're right that we have yet to see a consumer implementation in an AMD SoC. IIRC, there were rumors last year that failed to materialize.

The only reason I could see Nintendo going with x86 is for easier porting capability between the next console and X1/PS4. But for that to work, I would expect them to make drastic changes in how they approach and communicate with third party developers. It would require a culture change at Nintendo so I doubt that will happen.

What I think Nintendo should do is go with an existing 28nm SOC and not develop one of their own, something like a chip an existing Snapdragon line. It's probably already more powerful than what they would build, they're not on the hook for manufacturing as they will basically be buying a commodity part. You already have built-in Wifi (even LTE), camera, and all the peripheral connectivity they would ever need into the SOC. I think the net BOM would be lower than with a custom built component.

Snapdragon, I think, is a very good possibility for their QOL platform, where graphics will not matter. As far as their next traditional handheld/console go, judging by Iwata's comments, I am inclined to believe they are sticking with PPC. PS4/X1 ports are the only reason to go x86, and I don't think being a "me-too" console is on their agenda. Rather, sticking with PPC will allow them to optimize their OS over the lifespan of Wii U and also makes it so that VC and eShop games carry over. After their last failed console (Gamecube), Nintendo decided that BC was a priority, so there's a good chance they are thinking the same now. Combine that with making life easier on their own programmers and PPC seems an inevitability.

Now, how they are going to pull that off is beyond me. I'm assuming TSMC will manufacture since Renesas is pretty much out of the game. With AMD not shipping 20nm GPUs this year (and 20nm would appear to be vital if Nintendo is going to stick GCN cores in the handheld), Q2 2016 seems the earliest it will happen. Next year is QOL, after all.

I wonder if Nintendo would consider releasing their next console and handheld in the same year.

 

[Deleted member 11852]

x86-64 and GCN from AMD would be for instruction parity with the other consoles and PCs.

I like to reflect on Mark Cerny's interview with Eurogamer:

Digital Foundry: I guess from there we move to AMD as the partner of choice for creating PlayStation 4. Was it the research and products along the lines of the APUs that most excited you?

Mark Cerny: This probably doesn't quite answer your question, but there are so many issues involved in working with a vendor. The business relationship is very important. The timelines are very important, because you might be working with the brightest people in the business but if their product doesn't come out in the specific year that you need it, you can't work with them. There were a tremendous number of considerations and the choice of AMD came out of that.

It sounds like the 'choice' of x86 was less a decision and more the only option given their timeline and budget.

 

[Gradthrawn]

I like to reflect on Mark Cerny's interview with Eurogamer:



It sounds like the 'choice' of x86 was less a decision and more the only option given their timeline and budget.

From that very interview:

Digital Foundry: You were talking about the move to x86 as a viable console architecture. Microsoft tried it on the original Xbox with variable results, and it was pretty much dismissed after that. Can you talk more in-depth about the result of your research?

Mark Cerny: What I was looking at on my holiday was just the instruction set of the CPUs and it really seemed to me that things had been changing recently and maybe we were on the verge of a transformation in the x86 architecture and how it would be used. That didn't mean that we would necessarily use an x86 architecture - all that happened as a result of that was that I wondered if we could add it to our list of options. Then, when our hardware project kicked off for real in 2008, I and the ICE team and a number of other technology people within Sony did a very detailed analysis of x86 and how code would work on that versus other processors.

So we looked at the kinds of things we'd been doing on SPUs, we looked at general purpose code. We looked at writing in assembly for that platform versus others. We looked at compiled code and did about 15 different presentations showing off various aspects of the x86 architecture, and our first big meeting on PlayStation 4 was with the first-party teams. We spent most of the day going through our learnings on x86, we'd come to the conclusion that it was useful in a console but we had to ask them if they thought it could be useful in a console.

Which is then followed by the statement you quoted above. Based on that, and similar information in the Gamasutra interview it seems to me that x86 had been chosen before discussions with vendors even began. As such, it would be choosing AMD as the vendor that was "less a choice" and more a result of various business constraints. However, that's just my interpretation of the timeline and I don't think that's clearly spelled out anywhere.

 

[Grall]

Cross-platform.

Compilers can target different CPU architectures, it's not a dealbreaker by any stretch. It just requires nintendo actually making an effort this time on their developer tools. Multiplatform titles had it way worse in the past when consoles differed wildly in hardware architectures and capabilities and these issues were still overcome.

Unless Nintendo's strategy is to be the second console, and not a direct competitor for PS4/XB1, being easy for developers is very important.

There were no desktop class ARM cores back when 4bone was being developed, so X86 in their case was mostly a foregone conclusion, but as there are now there are alternative ways to gain similar performance without the baggage of x86. Likewise, PowerVR Rogue GPU cores offer DX11 feature-set as well as highly scaleable performance levels that surely reach current console levels if one were so inclined. Likewise, these options did not exist back when 4bone were still under development.

Alternatively, go with PPC or MIPS and awkward in-house tools and costly ports to a smaller userbase scaring away already apprehensive publishers

I'm fully convinced the current, incompetent Nintendo leadership could manage to produce awkward in-house tools even with x86 as a base. What's needed is change in nintendo leadership, not necessarily change in nintendo hardware architecture.

 

[Shifty Geezer]

...Nintendo decided that BC was a priority, so there's a good chance they are thinking the same now.

If Nintendo are going to wear the concrete boots of legacy HW designs just to satisfy their existing 5-6 million customers for BC, they deserve to drown in the sea of console competition. Their legacy since GC hasn't done them well and they shouldn't be dragging their heals with it. They need a future platform that'll be forwards compatible with what they intend to do, instead of burying their head in their halcyon past.

Wii U games aren't going to be that important on their new machine (and possibly won't even be possible if it doesn't have a Wuublet), while Wii and older can be emulated.

 

[Deleted member 11852]

From that very interview:.

Brief as O/T, you should also look at Cerny's speech at GameLabs in June last year. There were "two possibilities": PowerPC or 80x86. First party devs didn't want x86 but Cerny states "the potential design space of PlayStation 4 was going to be pretty limited if they couldn't use 80x86." Sorry, I don't have time to transcribe it all, find the full thing on YouTube, the CPU choice talk for PS4 starts 30 minutes in.

 

[Gradthrawn]

Brief as O/T, you should also look at Cerny's speech at GameLabs in June last year. There were "two possibilities": PowerPC or 80x86. First party devs didn't want x86 but Cerny states "the potential design space of PlayStation 4 was going to be pretty limited if they couldn't use 80x86." Sorry, I don't have time to transcribe it all, find the full thing on YouTube, the CPU choice talk for PS4 starts 30 minutes in.

Thanks, I'll check it out.

 

[Infinisearch]

I've had some thoughts on Nintendo lately so I guess I'll share them. First off I don't think they should get out of the console market, in fact I think they should do something in the interim to hold them over until next-gen.

Presenting the Wii-U+
1. 4GB of ram 3 dedicated to the game.
2. Replace all USB 2.0 port with USB 3.0 or 3.1 ports or Front 2 USB 3/3.1 Rear 2 eSATA.
3. Stick 2 SPU like cores on the CPU die.
4. If any or all of the above are possible (Thermals in small case, different process node) they will all need more power, increase power supply accordingly.

Reception
1. People who already bought the Wii-U will probably be pissed, but some might understand given the situation.
2. People who might have been interested in the Wii-U but didn't buy because the apparent lack of value (relative hardware power) compared to the then upcoming XB1 and PS4, might just buy it now.
3. Hardware Nuts still won't be happy.
4. People who will play a wait and see game to see if the upped specs will garner more first party developers for platform exclusives, and better chances at cross platform titles with XB1 and PS4.
5. Developers - Hopefully the more RAM and CPU boost will attract more developers to the platform.

Oh and the USB/eSATA upgrades are essential for faster external storage, ie the potential for faster asset streaming.

I really don't know what they should do but I can't really see them doing nothing.

 

[Grall]

They can't release a "wuu+", it would mean starting over from scratch - AGAIN - in the installed base department, how do you attract 3rd parties to something like that after only a year and a half-ish of launching your previous console?

Also, your specs are very weird. USB3 ports and eSATA, what the hell? So, you buy a new console - on top of the previous one you may just have bought - and then you have to buy an USB3 or eSATA enclosure also...? No, this'll never work. Consoles depend on being fixed, reliable, dependable systems that stay the same, with dependable performance characteristics and specs for the full lifetime of the console. That's why Sega fared so poorly starting in their later megadrive/saturn years; too many changes and add-ons.

Nintendo HAS to do nothing, for several more years. There's no other choice for them unfortunately.

 

[Megadrive1988]

After the the mainline console Zelda releases on Wii U, probably fall / holiday 2015, the console is going to be pretty much dead, barring an unprecedented turnaround this year, greater than what we saw with 3DS following its major price cut and release of important games.

In early 2016 the 3DS will be well past its prime, the sun will be setting on its life. The window for Nintendo's next gen platform(s) would be late 2016 or 2017 which would be somewhere between the launch of PS4 / XB1 in 2013 and the future consoles that are just now starting to be thought of at Sony and Microsoft, probably due by 2020 or perhaps a year earlier.

I'm only guessing, but I think the best, most affordable chip manufacturing node with the most capacity that will be available to Nintendo in 2016 would be 20nm.

 

[Goodtwin]

After the the mainline console Zelda releases on Wii U, probably fall / holiday 2015, the console is going to be pretty much dead, barring an unprecedented turnaround this year, greater than what we saw with 3DS following its major price cut and release of important games.

In early 2016 the 3DS will be well past its prime, the sun will be setting on its life. The window for Nintendo's next gen platform(s) would be late 2016 or 2017 which would be somewhere between the launch of PS4 / XB1 in 2013 and the future consoles that are just now starting to be thought of at Sony and Microsoft, probably due by 2020 or perhaps a year earlier.

I'm only guessing, but I think the best, most affordable chip manufacturing node with the most capacity that will be available to Nintendo in 2016 would be 20nm.

That's going to be a choice by Nintendo. They are forced to make a decision right now, do they go all in and fill their internal development pipeline with Wii U games, or do they chalk up the Wii U as a loss, and start preparing for a new hardware launch. Nintendo has enough internal resources that they could support the Wii U quite a while on their own. With that said, I think 2015 will look a lot like 2011 did for Wii, a new Zelda game on the list but very little else. Even though I think the Wii U will end up being a commercial failure, I still enjoy mine. The exclusives are enough to keep me happy with the hardware purchase.

 

[lwill]

I will step aside from talking about third-party relationship-building strategies, and will focus on what we may see in Nintendo's next-gen systems. From what NIntendo has been saying, I would expect them to make an ARM-based big mini-console (or small console.. which is sort of like what the Wii-U is). I believe it would be something like a Kindle FireTV device that was stripped and modified to be primary a gaming machine, and with a higher price and size budget ($200-250 range instead of $100) Due to the higher price and assuming that another Nintendo console will not be release until a few more years, I would expect the ARM-based SoC to readily surpass the Wii U, though I don't think it would have a chance to compete with XBO/PS4 in raw power.

The major advantage of this architecture is that it is not too dissimilar to familiar portable-tech, and it can be very similar to their portable successor. That will save Nintendo from extremely alpha dev-kits, enforce system unification, and that it would be easier to roll out an "upgrade" of the hardware as ARM tech advances further. However, I don't think such an architecture will allow them to completely "absorb" the Wii U architecture. 100% BC may have to be sacrificed so that Nintendo will have smoother transitions for the future.

As for the 3DS successor, I would expect it to be to the PSVITA in power as to how the 3DS is to the PSP. Even if Nintendo wants the portable to start at a considerably lower price than the 3DS ($100-150 range), the advancement in handheld tech will allow it to have a generation leap over the 3DS. There may be some issues with the battery life, though, especially if they want to enable 3D again. It will be interesting to see how that will turn out.

I know some posters believe that Nintendo will do more to absorb Wii-U's architecture, but don't see that happening to that extent at this time unless the portable successor is basically a portable Wii U. That route could allow them to make a much more powerful next-gen console, but I don't think that should be priority when you consider the task on making a powerpc/radeon-based portable.

~lwill

 

[Fourth Storm]

Does anyone here have any thoughts on where eDRAM is heading? Is there any manufacturer who would be able to provide it in another couple years or so? Seems like only IBM is really scaling theirs down to smaller nodes at the moment.