• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Nintendo NX rumored to use Nvidia's Pascal GPU architecture

Status
Not open for further replies.

MacTag

Banned
I think that if Nintendo is aiming for a cheap device, they will rather under clock it that use a fan.
The X1 devkits already use a fan though. So either the devkits are just temporarily using older overclocked chips (necessitating active cooling) or they plan to have a fan in the final units for cooling.
 

Eolz

Member
Question for those with more technological knowledge:

Is it possible for them to come out with the base handheld and powerless dock now, but then later supplement or replace the dock with something that adds power to the TV performance?

That way, the handheld stays the same power always but the TV performance can be increased?

It's possible in theory.
 

Akki

Member
So, isn't it logical for Nintendo to bring a powerful handheld first before a new console? That even can run unreal4 engine. 3ds has been released before wiiu.

This handheld could maybe run ports easily with a lower res.



Beside the handheld. Wiiu is 4y old. Wouldn't they better wait until next gen ps5 comes to release their new console?


People are disappointed because we are getting a handheld before a console, me too. But looking at the past, it's always have been like that.

Once more there is no reason to make a stand alone console because the hybrid is the console replacement. The beauty of this hybrid concept is that 3rd parties are more or less forced to deliver a HD version which massivly increases the software output compared to the Wii U (assuming there are different power levels).

From a business perspective a console has little chance to succeed compared to a hybrid or handheld. Let´s say Nintendo makes a 4TF machine at 400$ (rumored Neo numbers), while the handheld costs 200-250$. Obviously the higher cost lead to a higher entry barrier and most customer will chose the handheld.

Another problem is the big power difference between the console and the handheld. The bigger the gap the more ressources are required which increase the development time and lead to less software output. Just look at the PS4 first party output, most games require at least 3 years but Nintendo can just take all their Wii U projects and put them on the NX with little effort.

The sales will be heavily skewed towards the handheld. The japanese 3rd parties will put their games on the handheld (higher installbase) and avoid the console (see Wii U). They are not forced to develop a console version compared to the hybrid. Western 3rd parties won´t care at all because the audience is already non existent (but a strong hybrid + big installbase has at least a little chance to move AAA western software at a low attach rate). This will lead to less sales because there are fewer games and therefore Nintendo will sell few software on the Console even if they can increase their output compared to the Wii U thanks to the shared development. Most customors will buy the handheld because this way they´ll get the same first party games plus 3rd party games.

In the end we are looking at the same exact 3ds/Wii U situation. Some Wii U titles are heavily restricted or done cost efficiently because the sales potential is low. Just look at SF, Paper Mario, Mario Tennis or the AC game. I doubt Zelda:BotW would have been made as a Wii U exclusive. Nintendo will ask themselves if it´s worthwhile to develop expensive HD games if they can make more money at less risk and investment. The answer is obviously no.

Don´t expect a new console anytime soon. It makes far more sense to upgrade the hybrid after 2 or 3 years to milk their fanbase (like PS4 Neo or the new 3ds).
 

Malus

Member
Übermatik;212557503 said:
Yeah, I think people are misconstruing that information quite a bit... WSJ were probably talking in terms of yield/power consumption or something.

I was always skeptical of the "cutting edge" stuff, but I dunno why they'd be talking about yield/power consumption based on a demo they saw running on pc.
 

Rodin

Member
I don't think fan is an option from a cost perspective. Too many moving parts and possible overheating an damage if something goes wrong.

Nintendo more likely just be happy with whatever power they can get without a fan, as long as it surpasses wii u graphics and it can use modern engines.

Yes i can see this happening, but again, why describing the NX as something that blows the Wii U out of the water and it's somewhat "close" to the One? If it was "only" noticeably more powerful than the Wii U but still far from the Xbone (the 256-384 gflops ballpark we're reading here), i think the wording would've been different, not to mention the other reports we've heard.

Everything points out to a device noticeably less powerful than current gen consoles, but not closer to the Wii U than to those. Or at least that's how i'm reading it, which is why i'm looking for alternative explanations about what's going on.

Oh sorry, I misunderstood. It's still strange that Eurogamer and lherre (I'm gonna assume lowercase L, sorry if I'm wrong) haven't heard anything about varying power levels or resolutions if what you're saying is the case.

Maybe I'm fixating on that too much though.

No i think you're right, although lherre said that the fact that he didn't hear anything about that doesn't necessarily mean that it's not the case (or that this is something that it's already been finalized). DF not reporting about it is more "suspicious", in the sense that it's a hint towards the fact that it's not gonna happen, but who knows.

That's different than high-end PC.
Was this part of the WSJ report?
The development kit indicates that the system will have “incredibly powerful” hardware. Supposedly one demo included with the kit is currently impossible to run at 60fps on a current generation CPU and a “nearly top-of-the-line” graphics card.

Question for those with more technological knowledge:

Is it possible for them to come out with the base handheld and powerless dock now, but then later supplement or replace the dock with something that adds power to the TV performance?

That way, the handheld stays the same power always but the TV performance can be increased?
Isn't this basically the SCD patent?

Do you consider the ps4 cutting edge and industry leading as far as its tech as of today?

No, but this Pascal-based Tegra NX would be in the mobile space.
 
In my original predictions in the original NX speculation thread I actually stated that I expected a handheld first with a home console soon to follow. Threw out that idea with SMD64's leak.
I wonder if that's what's going to happen. Would be weird to make a NintendOS with systems acting "like brothers" with only one handheld. Though the TV connectivity on this new device is suspect.
The rumored specs are much higher than I was expecting with the handheld, so I'm happy either way. Will be interesting to hear just what's left to announce
 

Plinko

Wildcard berths that can't beat teams without a winning record should have homefield advantage
Isn't this basically the SCD patent?

Yes, pretty much. Guess I was wondering more what kind of power boosts we could see with this. My guess is this is what Nintendo has in mind, and we see a SCD sooner than later.

Initial rumors were that the "handheld" would come first with the "console" coming in a year or so. Perhaps the "console" in the rumor was an SCD to upgrade?
 
pascal_powercurve.png

.
 

Rodin

Member
In my original predictions in the original NX speculation thread I actually stated that I expected a handheld first with a home console soon to follow. Threw out that idea with SMD64's leak.
I wonder if that's what's going to happen. Would be weird to make a NintendOS with systems acting "like brothers" with only one handheld. Though the TV connectivity on this new device is suspect.
Maybe there is a standalone home console coming out in like 2019, but until then the NX is still more powerful than the Wii U, so they can use it as a home console too. If and when this hypotetical new home comes out, being able to connect the "hybrid" to the TV will still be a neat option, especially for those who don't want the other machine.

Yeah they did more then went out of their way to say thi is a powerful console that will outdue what we have in ps4/xb1. That was why people got their hopes up in what the console could deliver. Not just a fanboy dream without any merit.
There's actually an explanation for that statement that it's different from "this will be a high end console". Think of Quantum Break. It's a Xbox One game that runs like ass on high end PCs. In fact, it's impossible to run at 60fps with a "current gen CPU and a near top of the line GPU". There's no reason to assume that the demo they saw was perfectly optimized for that setup, hence why it didn't run great. Still, they said it was rather impressive, which isn't something they would say about the "Wii U+" graphics suggested here.

Yes, pretty much. Guess I was wondering more what kind of power boosts we could see with this. My guess is this is what Nintendo has in mind, and we see a SCD sooner than later.

Initial rumors were that the "handheld" would come first with the "console" coming in a year or so. Perhaps the "console" in the rumor was an SCD to upgrade?
Maybe, yeah. I don't think we can really discard any option about how it will work until we actually see this thing.
 
Was this part of the WSJ report?

No, that was some guy on a forum.

WSJ's Mochizuki did say this in an email last fall:

Several people who said who have seen a demo said what they saw is impossible to run on a computer without a “industry-leading” or “cutting-edge” chips. Cutting-edge in what way, they refused to elaborate.

...but as he himself acknowledges, that's vague, and his subsequent reporting from last week explicitly calls it a hybrid, so yeah.
 

Peterc

Member
The thing is that the latest info is no console at all. This is the new handheld and the new console.
Woukdn't mind at all if it was just the handheld, but the wiiu support is dying too much for that. They would ride out with some games like they did with the wii if another console was coming later.

I'm with you.

Because the wiiu is dying they would use this handheld also as console. Because it could run ported games from ps4 probably. That way they could have a big library and forget the wiiu. So yes it would be some kind of replacement too until next gen comes.


It would be the only way to make profit and get back in the race with their competitors(but still doing it on their way and still get 3rd party support).
 

Vena

Member
The X1 devkits already use a fan though. So either the devkits are just temporarily using older overclocked chips (necessitating active cooling) or they plan to have a fan in the final units for cooling.

The dev kits are using a standard X1 in full (or more) mode to run code + overhead.

Any one who's undershooting whatever the X1 can do stock (minus some-% on the overhead costs) is playing against the odds even if this "lolNintendo" because no one is shipping out dev kits like this that they plan to then severely undercut with the next generation of chip (no less) a couple of months later with the retail product. The fact that the X1 is actively cooled is not, however, an "oh shit" moment since we don't know what that fully entails, it doesn't tell us if that is necessitated by the code (optimized or not) or by the overhead or by both, and so on.

The games being developed simply would not work in the scenario where "lolNintendo" is put into affect and we get something so drastically below even the stock X1. This goes for third and first party titles, the internal dev teams don't have some magical other unit or were somehow told "oh hey, you're getting actually Y".

Even if you assume some nonsensical 50% overhead and unoptimized spaghetti code, that's still nowhere near the nonsense that has been floating around in the last few pages. And even that mess of a situation would still double the WiiU. In a handheld.
 

Rodin

Member
No. That came from a forum user on a different website.

You're right, it was a mod or admin of that Nintendo forum i think. This was the actual follow up from WSJ about the "industry leading" stuff

Hello David,

The quick answer is “I don’t know.” Sorry!

-Several people who said who have seen a demo said what they saw is impossible to run on a computer without a “industry-leading” or “cutting-edge” chips. Cutting-edge in what way, they refused to elaborate.
-And an important thing to remember, probably you know well already, is that chip specs won’t be finalized until much closer date to the release.

hope this helps, and plz ask me follow-up questions if you have.

thanks,

mochi
 

Eolz

Member
I'm with you.

Because the wiiu is dying they would use this handheld also as console. Because it could run ported games from ps4 probably. That way they could have a big library and forget the wiiu. So yes it would be some kind of replacement too until next gen comes.


It would be the only way to make profit and get back in the race with their competitors(but still doing it on their way and still get 3rd party support).

Maybe. Again, there's always the third pillar excuse (they can say later it was a successful experiment but consoles are coming back as well for those wanting more).

They really need to be clear at the reveal and interviews afterwards...
 
That's different than high-end PC.

Without context, it's even possible that Mochizuki's sources saw something like the ProtoStar demo and mistook it for something that would require high-end PC/console hardware. If that sounds far-fetched, it really isn't - IGN ran a report in 2010 that cited developer sources likening 3DS' performance to the PS3/360, presumably based on little more than its support for some modern shader effects like normal mapping and self-shadowing.
 

MuchoMalo

Banned
A53 is maybe around half the A57s performance, but it's proportionally much smaller than half the die size, so 8 of those over 4 A57s can be done for die area saving. Ideally I'd want 4 A57s, but if they go with A53s 8 makes sense to me. they're trivially small to throw on a die. It's why many cheap phones are octa core while better ones often have less cores which can look odd to untrained consumers.

Losing 4 A53 cores would be nowhere near 1SM of die size.

Not dissimilar to having 8 Jaguar cores vs 4 or even 2 that provide the same performance, but those larger cores scale more in die area than they do in performance


http://images.anandtech.com/doci/7995/Screen Shot 2014-05-06 at 2.59.56 AM.png

What I mean is that having 8 cores is extreme overkill for a Wii U-level machine, even if those cores are A53s. 4 would make more sense with terrible specs like those. Granted, I'm extremely confident that there will be two SMs anyway, but a quad-core A53 @1.7GHz should be enough power for two Pascal SMs @500-600MHz. 8 cores for one SM is just pointless.

What is so hard about dev's optimizing for two different configurations? It could easily be a part of Nintendo's lot checks, ensure the game runs in undocked mode at acceptable performance. The Dev would just have to have a separate configuration for graphic settings, LOD's etc.. If it's a known quantity dev's can adapt.


As far as I know we don't know the clocks of the dev kits. And I believe the X1 under normal conditions does need a fan, though for retail products it isn't a significant one.

Read the post I was replying to. I believe that he was implying that devs simply don't know about the underclocked mode, and the system would simply dynamically underclock and lower the resolution of games in handheld mode. That's not possible.

lol

Why did this thread spiral so badly for like three pages? lherre's input was nothing outside of what we already knew. This thing is and never would be going to match PS4s or therein as it is a handheld.

you_are_putting_words_in_my_mouth.gif


Seriously, nobody was thinking this.
 

Peterc

Member
Once more there is no reason to make a stand alone console because the hybrid is the console replacement. The beauty of this hybrid concept is that 3rd parties are more or less forced to deliver a HD version which massivly increases the software output compared to the Wii U (assuming there are different power levels).

From a business perspective a console has little chance to succeed compared to a hybrid or handheld. Let´s say Nintendo makes a 4TF machine at 400$ (rumored Neo numbers), while the handheld costs 200-250$. Obviously the higher cost lead to a higher entry barrier and most customer will chose the handheld.

Another problem is the big power difference between the console and the handheld. The bigger the gap the more ressources are required which increase the development time and leads to less software output. Just look at the PS4 first party output, most games require at least 3 years but Nintendo can just take all their Wii U projects and put them on the NX with little effort.

The sales will be heavily skewed towards the handheld. The japanese 3rd parties will put their games on the handheld (higher installbase) and avoid the console (see Wii U). They are not forced to develop a console version compared to the hybrid. Western 3rd parties won´t care at all because the audience is already non existent (but a strong hybrid + big installbase has at least a little chance to move AAA western software at a low attach rate). This will lead to less sales because there are fewer games and therefore Nintendo will sell few software on the Console even if they can increase their output compared to the Wii U thanks to the shared development. Most customors will buy the handheld because this way they´ll get the same first party games plus 3rd party games.

In the end we are looking at the same exact 3ds/Wii U situation. Some Wii U titles are heavily restricted or done cost efficiently because the sales potential is low. Just look at SF, Paper Mario, Mario Tennis or the AC game. I doubt Zelda:BotW would have been made as a Wii U exclusive. Nintendo will ask themselves if it´s worthwhile to develop expensive HD games if they can make more money at less risk and investment. The answer is obviously no.

Don´t expect a new console anytime soon. It makes far more sense to upgrade the hybrid after 2 or 3 years to milk their fanbase (like PS4 Neo or the new 3ds).

I can follow you and also agree for most parts.


But, I assume that sony and ms is also bringing the ps4 neo and xs to keep the console longer alive. Sony told with the ps3 that they wanted to give it a lifespan of 7years. I think they want to expand the lifetime with ps4 neo for even 4y+ if they can. Games are getting way to expensive. 100m$ is the max budget and they have reached that point.


Knowing this. The handheld could have a long lifetime, this is what Nintendo also expect for the nx.

But what happens after 4y's, i don't think they could survive another 3y extra when next gen comes using their handheld.


The only solution maybe are the extra power base station and maybe some upgrade. But i doubt it would get many support.

But the companies that building consoles thinking of upgrading their systems like apple/android does are wrong. I don't believe people would upgrading their console,it's not as important as a smartphone they carry everyday.


Beside all this, we don't know what would happen with vr. Maybe it replace the traditional way of playing games.
 

Rodin

Member
Read the post I was replying to. I believe that he was implying that devs simply don't know about the underclocked mode, and the system would simply dynamically underclock and lower the resolution of games in handheld mode. That's not possible.

you_are_putting_words_in_my_mouth.gif


;)

Read my other post, i wasn't saying that, but i didn't explain that part very well so i can see why people would misunderstand it.
 
Maybe there is a standalone home console coming out in like 2019, but until then the NX is still more powerful than the Wii U, so they can use it as a home console too. If and when this hypotetical new home comes out, being able to connect the "hybrid" to the TV will still be a neat option, especially for those who don't want the other machine.
Yeah, that's one way to look at it. They do need to replace Wii U and 3DS soon so a hybrid like system could do the trick while they work on/consider a console for 2018/19
It's a reasonable thing to question, at least Nintendo thinking about future systems, it wouldn't be worth mention if it was just one form factor.
Yep, while looking back Iwata didn't discard the idea of a hybrid, it does indicate that they have the capability to create multiple devices. Part of Iwata's reasoning was to reassure that they weren't leaving the hardware business but that they were more passionate than ever. Deciding to only make one device with their OS would go against that.
 

NateDrake

Member
Without context, it's even possible that Mochizuki's sources saw something like the ProtoStar demo and mistook it for something that would require high-end PC/console hardware. If that sounds far-fetched, it really isn't - IGN ran a report in 2010 that cited developer sources likening 3DS' performance to the PS3/360, presumably based on little more than its support for some modern shader effects like normal mapping and self-shadowing.

Yes. Yes. Exactly.
 
The dev kits are using a standard X1 in full (or more) mode to run code + overhead.

Any one who's undershooting whatever the X1 can do stock (minus some-% on the overhead costs) is playing against the odds even if this "lolNintendo" because no one is shipping out dev kits like this that they plan to then severely undercut with the next generation of chip (no less) a couple of months later with the retail product. The fact that the X1 is actively cooled is not, however, an "oh shit" moment since we don't know what that fully entails, it doesn't tell us if that is necessitated by the code (optimized or not) or by the overhead or by both, and so on.

The games being developed simply would not work in the scenario where "lolNintendo" is put into affect and we get something so drastically below even the stock X1. This goes for third and first party titles, the internal dev teams don't have some magical other unit or were somehow told "oh hey, you're getting actually Y".

Even if you assume some nonsensical 50% overhead and unoptimized spaghetti code, that's still nowhere near the nonsense that has been floating around in the last few pages. And even that mess of a situation would still double the WiiU. In a handheld.

What do you mean by this?
 

Mpl90

Two copies sold? That's not a bomb guys, stop trolling!!!
Interesting. So we have sources saying pascal and sources saying maxwell.

I suppose the "sources" Digitimes references to is the EuroGamer's article where they state it's X1 in the dev kits. Anyway, it's better to not take DigiTimes so seriously.
 

KingSnake

The Birthday Skeleton
What happened to this thread in the last pages? This is going insane, but not in the WUST way, in a mass-denial way. It's like seeing a apocalyptic cult or something.
 

Mpl90

Two copies sold? That's not a bomb guys, stop trolling!!!
What happened to this thread in the last pages? This is going insane, but not in the WUST way, in a mass-denial way. It's like seeing a apocalyptic cult or something.

It's...this hybrid...
 

Vena

Member
I did not understand the paragraph... are you saying that in the worst case scenario, the nx would be twice as powerful than wii u?

Yes. Thats the scenario where the X1's are being used at stock (active cooling for whatever reason) because they have 50% overhead on tools and spaghetti code. Basically 50% of the units capabilities are just used to negotiate bad code, observer tools, and so on, and only the remaining 50% is actually running a game.

That still give you a performance envelope of double the WiiU's.

And I highly doubt that that is the case.

The other point was that actively cooled or not, doesn't actually tell us much of anything other than for whatever task it is expected to currently do, it is doing it with too much heat. This could be because it has a lot of overhead (see above) or because the Pascal replacement is expected at similar performance levels but without needed active cooling. Or both.

Right now, ballparking at "whatever X1 can do at stock" is probably not a bad ballpark, give or take. But we have people going mining or shooting for the stars on nonsensical hyperbole trails.
 

MuchoMalo

Banned

Interesting. So we have sources saying pascal and sources saying maxwell.

Digitimes isn't reliable. They change their story every time new leaks or information pop up.

This is interesting. How reliable has digitimes been in the past?

Not at all. First it was production starting in H1 CY 2016 with estimated demand for 20 million units sold in 2016 but only meeting production for 12 million, then it was production delayed to Q1 2017 with production cut from 20 million in the first year to 10 million, and now it's production moved up to Q4 2016 and them hearing about Tegra X1 all of a sudden. Either NX is in development hell with initial plans to launch this summer, or they're full of crap. Someone's gonna make a thread eventually, though.
 
So what is the NX platform, and why did Iwata emphasize that NX represents a platform?Platform suggest that anything NX is handheld or home console. With Iwata passing could Nintendo direction have changed?

Is it possible Osiris info is from Japanese devs who would be much further along in the development process, and have gotten SCD kits already.
 

Thraktor

Member
So that would still be a 300+ Gflop device, which is about 2x Wii U, right? I would be fine with that, though I can also see where some people would be disappointed.

Edit: I realize these are just estimates

Yeah, my median estimate would be around 300 Gflops (FP32) or so, which, incidentally, is 50% more powerful than current high-end smartphones, while pushing a (likely) much lower resolution screen (and that's not taking into account architectural differences).

2W? Well, now I understand your CPU clock predictions. What screen size are you guessing with that prediction?

I'm really not sure if the CPU will even beat Wii U's across the board with a 1W budget.

You could easily beat Wii U's CPU with a 1W budget. That would give you either 8x A53 at 1.7GHz or 4x A72 at 1.2GHz. Or you could mix and match with 2x A72 at 1.4GHz plus 6x A53 at 1.1GHz. All three of those options would be closer to PS4's CPU than Wii U's.

In that case would it actually be feasible to get at least stock TX1 performance (512GFlops) in a tablet on battery mode with a Pascal based Tegra chip? That would be a pretty damn beefy tablet! Interesting idea...

You could do that in a regular handheld with 3 SMs and a ~3W TDP. It's very much at the upper end of what I'd consider feasible, but you wouldn't need a tablet form factor for it.


293726.gif
 
So what is the NX platform, and why did Iwata emphasize that NX represents a platform?Platform suggest that anything NX is handheld or home console. With Iwata passing could Nintendo direction have changed?

Is it possible Osiris info is from Japanese devs who would be much further along in the development process, and have gotten SCD kits already.

WSJ's reporter is Japan-based.
 

ethomaz

Banned
The bolded isn't true. In fact, pretty much the opposite is true, as in a tightly thermally constrained environment (i.e. a handheld) the marginal benefit to increased parallelism (i.e. more SMs) can be quite large.

To demonstrate, let's look at a power curve for Pascal I've put together. Unlike my previous power curves for A72 and A53 CPU clusters (which are based on solid real-world data from Anandtech and should be considered reasonably accurate), this is a much more rough approximation based on just four data points:

- TSMC's claims of "40% higher speed" and "60% power saving" over 20nm, each applied separately to the TX1's GPU drawing 1.5W at 500MHz (divided by 2 for 750mW per SM).
- Power draw readings from the GTX1080 before and after overclocking (full board power readings, minus GDDR5X, divided by number of SMs).

Obviously I'm extrapolating a lot from fairly poor data, but hopefully it should be in the right ballpark, and enough for our discussion in any case. (I should also note that this isn't strictly a measure of power draw for the SMs themselves, but rather a measure of the draw of an entire Pascal GPU "per SM", so including other components like ROPs, TMUs, etc., assuming they're always in roughly the same proportion to SMs). In any case, here's the power curve:

pascal_powercurve.png


The important thing to note is that, like virtually all IC power curves, it's not linear, and for a given increase in clock speed you require a much larger increase in power consumption to get you there. What this means is that you'll get better performance by using more SMs at a lower clock speed than fewer SMs at a higher clock speed.

Let's look at the clock speed (and raw floating point performance) that could be achieved with different numbers of SMs within the power constrains we might expect for a handheld GPU:

1x SM:

1000 mW - 780 MHz - 200 Gflops FP32 - 400 Gflops FP16
1500 mW - 915 MHz - 234 Gflops FP32 - 468 Gflops FP16
2000 mW - 1025 MHz - 262 Gflops FP32 - 525 Gflops FP16

2x SM:

1000 mW - 595 MHz - 305 Gflops FP32 - 609 Gflops FP16
1500 mW - 700 MHz - 358 Gflops FP32 - 717 Gflops FP16
2000 mW - 780 MHz - 400 Gflops FP32 - 800 Gflops FP16

3x SM:

1000 mW - 510 MHz - 392 Gflops FP32 - 783 Gflops FP16
1500 mW - 600 MHz - 461 Gflops FP32 - 922 Gflops FP16
2000 mW - 670 MHz - 515 Gflops FP32 - 1030 Gflops FP16

As you can see, a 3x SM configuration can achieve nearly the same performance with 1000mW that a 2x SM configuration can with twice that, and a full 50% more than a 1x SM config can manage with 2000mW at hand.

This isn't to say that I expect a 3x SM GPU in the NX, but there would certainly be a sizeable performance jump over 2x SMs if they decided to do so.
You can add 4 SMs because that is the number I think nVidia will use to Tegra X2.

Said that anything under 5000 mW is a option for NX.
 

LordOfChaos

Member
Yeah, my median estimate would be around 300 Gflops (FP32) or so, which, incidentally, is 50% more powerful than current high-end smartphones, while pushing a (likely) much lower resolution screen (and that's not taking into account architectural differences).

It's 50% more Gflops at any rate, I believe Imagination's PowerVR still has the efficiency thing down compared to anyone else and their 200Gflop GPUs usually edge out anyone elses 200Gflop GPUs.

Case in point, the highly impressive iPhone 6S GPU "only" weighs in at 115Gflops and gives everything else a run even a year after launch (it's slipping in the graphs now, but for something so old it still impresses, and does so with far less paper gflops than others...And without overheating and severe throttling in minutes...Apple knows how to pick 'em). The GS7 is closer to 265-300 depending on the model.
 

KingSnake

The Birthday Skeleton
So what is the NX platform, and why did Iwata emphasize that NX represents a platform?Platform suggest that anything NX is handheld or home console. With Iwata passing could Nintendo direction have changed?

It's very possible NX to be the basis for a platform going forward/ There could be other devices built on similar architecture further down the line. That doesn't mean that it will happen all in 2017. Could be 2-3 years before another device comes. Or there could be no other device yet in development. Maybe all depends on this NX device's success.

Or maybe Iwata used platform to not give any hint about what they're working on. Why would they hint that 2-3 years before release when they're very secretive even about the games they're working on a lot of the time.
 
If the device is a tablet as I am now on the side of, I understand that we can sacrifice die size limitations and give it some increased power draw. My question is, because Nintendo needs a bigger screen and maybe a bigger battery, doesn't the device end up being that much more expensive?
 
Status
Not open for further replies.
Top Bottom