Support NeoGAF

[Azoo]

Considering the 3DS is very likely to last throughout the next year (up to Q3 2017 I imagine), I don't believe the gimmick is going to be about any portable that features a shared library... yet.

That does leave me wondering what they're going to do, though. I don't think we're getting another Wii, but I also don't think scrolly shoulder buttons would be all they'd attempt.

 

[Graphics Horse]

It doesn't seem like Nintendo is keen on replacing the 3DS quite yet. Their comments about the larger installed base and being able to sell to it at the Investor Briefing make me think we're getting a console first. Your suggestion would be a very interesting device to release late in 2017, though, especially if the shared library comes to fruition and people can play Wii U levels of fidelity for Smash Brothers and Breath of the Wild on their portable.

The 3DS brings in lots of money, so I am sure they are not keen to have sales dry up by November. Ideally whatever system they bring out it will be healthy for (at least) one existing system to still be selling alongside. That said, I don't expect much software support into 2017, is Ever Oasis the only announcement? Edit: Not forgetting Kawashima’s Devilish Brain Training TBA.

 

[lwilliams3]

By my estimates, it would cost Nintendo only around $12-$13 for a 16nm Nvidia chip for a March 2017 launch with the following:

• 4-core A53 CPU
• Pascal GPU w/ 2 SMs (256 "cores")
• 64-bit LPDDR4 interface

You would be able to clock the CPU at 1GHz and the GPU at 400MHz for a total power draw of about 1.85W, which would give you over 200 Gflops of raw GPU power (with a far more modern GPU arch than Wii U). This is well within the expected power draw for a handheld like the 3DS or Vita, and should be capable of providing pretty good battery life (depending on other components, of course). The cost is only slightly more than the $10 the 3DS SoC was expected to cost at release.

Granted, my calculations have a large margin of error, but it's certainly within the realm of possibility that Nintendo could release an affordable handheld in early 2017 that's more powerful than the Wii U.

Thank you for the info. I missed your previous post about this. I knew that Nintendo could make a portable that could output graphics beyond the Wii U on a small screen, but I didn't think it would be reasonable to straight up surpass its specs (except for the RAM and CPU which will take little effort.)

 

[Dark Cloud]

Man my hype is so high for the NX. And to think we could see its full reveal in a few weeks

I can't wait to see what they come up with.

 

[diaspora]

I don't know why people keep asserting nvidia, no console maker is going to use them =\

 

[10k]

Man my hype is so high for the NX. And to think we could see its full reveal in a few weeks

I can't wait to see what they come up with.

Scott Moffitt said this fall though do far away

I don't know why people keep asserting nvidia, no console maker is going to use them =

You know why.

 

[G.ZZZ]

By my estimates, it would cost Nintendo only around $12-$13 for a 16nm Nvidia chip for a March 2017 launch with the following:

• 4-core A53 CPU
• Pascal GPU w/ 2 SMs (256 "cores")
• 64-bit LPDDR4 interface

You would be able to clock the CPU at 1GHz and the GPU at 400MHz for a total power draw of about 1.85W, which would give you over 200 Gflops of raw GPU power (with a far more modern GPU arch than Wii U). This is well within the expected power draw for a handheld like the 3DS or Vita, and should be capable of providing pretty good battery life (depending on other components, of course). The cost is only slightly more than the $10 the 3DS SoC was expected to cost at release.

Granted, my calculations have a large margin of error, but it's certainly within the realm of possibility that Nintendo could release an affordable handheld in early 2017 that's more powerful than the Wii U.

Well, i was amply proven wrong. What make me doubt is... would nintendo be open on selling this thing with said margins? I mean, they sell , right now, the 3DS at 199 $. I guess that's like a 90% margin jeez.

I really liked that other post of yours. I don't think however that Nintendo has to run the exact same chip only with different clocks. Even same-family GPUs would make scaling of the games between mobile->home console fairly trivial , but who know, maybe the advantages of ordering just mass of 1 single chipset (and then purpose them depending on the destination) is enough economic incentive for them to do so. I just hope we don't get worse than XB1 specs cause i'd like to have games render at the native res of my TV and not upscale like XB1 (this may not be a concern at all if only nintendo develop on it, but if we're getting even just a few ports, i'd like to get the native 1080 res)

EDIT: oh i see you're skeptical too about using the exact same SoC and did the calculation that way just because.
Yeah nothing much to add here.

 

[MuchoMalo]

Well, i was amply proven wrong. What make me doubt is... would nintendo be open on selling this thing with said margins? I mean, they sell , right now, the 3DS at 199 $. I guess that's like a 90% margin jeez.

I really liked that other post of yours. I don't think however that Nintendo has to run the exact same chip only with different clocks. Even same-family GPUs would make scaling of the games between mobile->home console fairly trivial , but who know, maybe the advantages of ordering just mass of 1 single chipset (and then purpose them depending on the destination) is enough economic incentive for them to do so. I just hope we don't get worse than XB1 specs cause i'd like to have games render at the native res of my TV and not upscale like XB1 (this may not be a concern at all if only nintendo develop on it, but if we're getting even just a few ports, i'd like to get the native 1080 res)

The SoC is only one part of the device, and not even the most expensive part. You also forgot about the retailer cut. Nintendo's margin can't be more than 50%, and is likely closer to 30%.

The advantage to making only one chip is drastically reduced costs over time. With that, both the console and the handheld would be no more than $200. However, this would result in a console significantly weaker than Xbone since a handheld's SoC can't be much larger than ~100mm2, which is a third of Xbone's. (Before I get pounced, I'm just explaining something here.)

 

[DizzyCrow]

By my estimates, it would cost Nintendo only around $12-$13 for a 16nm Nvidia chip for a March 2017 launch with the following:

• 4-core A53 CPU
• Pascal GPU w/ 2 SMs (256 "cores")
• 64-bit LPDDR4 interface

You would be able to clock the CPU at 1GHz and the GPU at 400MHz for a total power draw of about 1.85W, which would give you over 200 Gflops of raw GPU power (with a far more modern GPU arch than Wii U). This is well within the expected power draw for a handheld like the 3DS or Vita, and should be capable of providing pretty good battery life (depending on other components, of course). The cost is only slightly more than the $10 the 3DS SoC was expected to cost at release.

Granted, my calculations have a large margin of error, but it's certainly within the realm of possibility that Nintendo could release an affordable handheld in early 2017 that's more powerful than the Wii U.

Damn, I would trade any "innovation" for the ability to play all games on both console and handheld (or at least all handheld games on the console) just a straight up game machine.

 

[Dark Cloud]

Scott Moffitt said this fall though do far away


You know why.

Lol! I meant to say a few months!

 

[G.ZZZ]

The SoC is only one part of the device, and not even the most expensive part. You also forgot about the retailer cut. Nintendo's margin can't be more than 50%, and is likely closer to 30%.

The advantage to making only one chip is drastically reduced costs over time. With that, both the console and the handheld would be no more than $200. However, this would result in a console significantly weaker than Xbone since a handheld's SoC can't be much larger than ~100mm2, which is a third of Xbone's. (Before I get pounced, I'm just explaining something here.)

As for Thraktor analysis:

On the home console front you might be looking at, say 8-core A72, 6 SMs and 256-bit LPDDR4 interface, which would come to $46.32.

For over 1.5+ NV Tflops a cost of about 45-50$ (current WiiU cost 100 because Nintendo went for a particular custom design and an extremely old CPU chip).

I'm no expert in knowing how much that is actually doable.

 

[Gabe3208]

Even most Nintendo fans aren't with you on this.My favorite Nintendo games from the last decade didn't need or significantly benefit from the special features of the platforms they were released on. And it's not just me, look at which Nintendo games from the last decade are most critically acclaimed. Nintendo couldn't make a compelling case for motion controls or tablet controls. And while twin screens and 3D have been more successful, they haven't become industry standards.

Arguably the most beloved Nintendo console is the SNES, which also happens to be the least innovative home console Nintendo has put out. What it did have was a major graphics and sound boost over its predecessor, a better controller, and superb third-party support.

What makes Nintendo special is the quality of their games, not that their hardware has unique capabilities that even they don't quite know what to do with.

And that's perfectly fine, I didn't say I was speaking for every Nintendo fan. I was expressing how I felt about them myself.

And maybe the quality of their games is what makes them special to you, but for me, it's much more than that. To actively try to differentiate yourself from all the others, it's a very special thing to continue to do today, even when you know how much of a risk it is. I love new toys, new concepts. I will still buy a PlayStation and/or Microsoft console because of 3rd party games that don't appear on Nintendo consoles, but I'd be lying to myself if I said that the level of excitement I have for those platforms is the same as with Nintendo's consoles.

You can go to a theme park that has the tallest and fastest roller coasters/rides in the world, but there is no experience like being in Disney World with their "not so crazy thrilling" rides. The atmosphere/environment just makes you forget about all that and just lets you relax and enjoy what's there. It's the same with Nintendo, to me, they're untouchable in that magical sense.

 

[Jubenhimer]

And that's perfectly fine, I didn't say I was speaking for every Nintendo fan. I was expressing how I felt about them myself.

And maybe the quality of their games is what makes them special to you, but for me, it's much more than that. To actively try to differentiate yourself from all the others, it's a very special thing to continue to do today, even when you know how much of a risk it is. I love new toys, new concepts. I will still buy a PlayStation and/or Microsoft console because of 3rd party games that don't appear on Nintendo consoles, but I'd be lying to myself if I said that the level of excitement I have for those platforms is the same as with Nintendo's consoles.

You can go to a theme park that has the tallest and fastest roller coasters/rides in the world, but there is no experience like being in Disney World with their "not so crazy thrilling" rides. The atmosphere/environment just makes you forget about all that and just lets you relax and enjoy what's there. It's the same with Nintendo, to me, they're untouchable in that magical sense.

I would also like to add that new hardware concepts can properly facilitate new types of games. Truama Center for example, wouldn't have been nearly as playable or unique without the DS' touchscreen. Elebits, couldn't have been done without the Wii Remote. Innovation comes from both hardware and software, not one or the other.

 

[MuchoMalo]

As for Thraktor analysis:



For over 1.5+ NV Tflops a cost of about 45-50$ (current WiiU cost 100 because Nintendo went for a particular custom design and an extremely old CPU chip).

I'm no expert in knowing how much that is actually doable.

It's doable in a console, but requires separate chips for the console and handheld. Of course, Thraktor is assuming a 28nm Maxwell chip here so I don't know how well the cost would really translate to Pascal, which I still believe that they're using. I can't see Maxwell because it would require nearly twice as much power as Wii U and I honestly don't think 16nmFF+ costs as much extra as Thraktor believes. Either way, if that analysis is correct and we throw in a card slot in place of an optical drive, even with a not-so-cheap gimmick we can come in at under $300 while outpacing PS4 handily... So, expect a $200 machine which can't touch Xbone but nets a healthy profit.

Edit: Well, this is before considering that the costs of 8GB LPDDR4 and the cost of storage (unless they skimp again), so it would actually still lose money at $300... Welp.

 

[MDave]

Which ever chips they use, I hope they go for the strategy of using something powerful but off the shelf, instead of low power, custom or legacy design (Wii U).

So the cost of the chips falls much faster, we get the power we want, the devs get the power they need, and Nintendo get the profits they want when the chips are much cheaper. The cost of chips will be different a year from now, and even more 2 years in.

 

[KirbKirb30]

Let us introduce, the Nintendo PiiU. It accurately reproduces the smells for what is on screen.

Comes packed with Poopy Mario Bros.

I absolutely should not have laughed at that but I did.

 

[Vanillalite]

Unveiling the NX would probably get functionally burried during the Holidays.

So reasonably speaking Nintendo has basically 4 months to get their shit in order for a reveal unless they miss their 2016 reveal year.

I know Miyamoto says they are trying to iron out some shit, but idk how much more leash they can get.

They also (probably) aren't gonna announce the same week as something else. So you could cross out the week of TGS and Gamescom. You could also cross out the week of the fall Apple event. Ditto for Sony if they decide to have their own standalone event for Neo.

So that that leaves them with like 3 months realistically.

July seems to be the best month in terms of nothing else to compete with in the news cycle, but also seems the least likely due to this new info from Miyamoto.

The close they get to fall the busier shit gets though.

 

[DizzyCrow]

What makes Nintendo special is the quality of their games, not that their hardware has unique capabilities that even they don't quite know what to do with.

Well said, looks like Nintendo is not capable to think that maybe, just maybe their amazing games are enough for most people and they don't need to keep gambling their future every time.

 

[Thraktor]

It doesn't seem like Nintendo is keen on replacing the 3DS quite yet. Their comments about the larger installed base and being able to sell to it at the Investor Briefing make me think we're getting a console first. Your suggestion would be a very interesting device to release late in 2017, though, especially if the shared library comes to fruition and people can play Wii U levels of fidelity for Smash Brothers and Breath of the Wild on their portable.

I agree, and I fully expect that what we're getting in March is a home console (particularly due to Nintendo specifically citing the existence of NX as a cause for the drop in their Wii U forecasts this FY), but my point was more that they probably could release a handheld capable of running Wii U games like Zelda:BOTW in early 2017 if they wanted to, without having to sacrifice battery life or charge an extortionate amount. Late 2017 seems a more sensible timeframe, though, particularly as 16nm matures and becomes more affordable.

It's doable in a console, but requires separate chips for the console and handheld. Of course, Thraktor is assuming a 28nm Maxwell chip here so I don't know how well the cost would really translate to Pascal, which I still believe that they're using. I can't see Maxwell because it would require nearly twice as much power as Wii U and I honestly don't think 16nmFF+ costs as much extra as Thraktor believes. Either way, if that analysis is correct and we throw in a card slot in place of an optical drive, even with a not-so-cheap gimmick we can come in at under $300 while outpacing PS4 handily... So, expect a $200 machine which can't touch Xbone but nets a healthy profit.

Edit: Well, this is before considering that the costs of 8GB LPDDR4 and the cost of storage (unless they skimp again), so it would actually still lose money at $300... Welp.

Nope, my calculations are all based on Pascal* on TSMC's 16nm process.

Besides, 8GB of LPDDR4 (or whatever other memory they were to use, sans HBM) wouldn't be prohibitively expensive for a March 2017 release, probably around $50. On the storage side, the 32GB of UFS memory used in the S7 cost just $7.25, and even if (a) there were no returns to scale on flash and (b) cost didn't come down at all in the year since the S7 release (both blatantly untrue) they could still get 128GB of (very fast) UFS flash for $29, less than the HDD cost in launch PS4s. More realistically, they could probably get 128GB for under $20 and 256GB for under $30. Without mandatory installs, and with USB3 external storage support, either of those would be perfectly satisfactory.

All in all, a $50 SoC, 8GB of LPDDR4 and 256GB of flash should be entirely doable for a $250-$300 console with no disc-drive and no expensive additional features.

*GP104-style Pascal with 128-core Maxwell SMs rather than GP100-style "true" Pascal with 64-core SMs.

 

[neoemonk]

I was thinking the 3DS was getting pretty long in the tooth until I realized the DS was out for seven years before the 3DS even came on the scene.

Whether or not the NX is going to be a combo console/handheld I'm looking forward to the eventual reveal.

 

[ecosse_011172]

There's a few Twitter "insiders" still camping that the NX will be between ps4 and neo but closer to neo.

I think there will be disappointment, I wonder if there is one Polaris hoaxer feeding all of them, 10k etc.