Nintendo Switch Dev Kit Stats Leaked? Cortex A57, 4GB RAM, 32GB Storage, Multi-Touch.

Status
Not open for further replies.
I don't know what kind of deal could Nvidia offer to Nintendo to go for 20nm when 20nm would be bad for them on so many levels. Everybody is running away from it. To cover the production costs in the long run and to make up for the lower battery life and/or lower performance it should be the deal of the decade, really.

And I posted already this in one of the threads, I don't remember now which ones, but TMSC is moving away very fast from 20nm production (according to their own financial results conferences starting from Q1 this year - I'm too lazy now to link it), so who would produce this?

But who knows, maybe Nintendo did get the deal of the decade after all.
 
I just don't believe nvidia could have convinced them to go with a process nobody else wanted. Nintendo is cheap but they aren't THAT cheap.
 
That's only a 5% difference.

TSMC claims a 1.1x density improvement over 20nm. It's enough, along with power and performance gains, to be worth pursuing. They've also improved their substrate technology to allow for a larger die in smaller form factors. Nintendo paid Kings ransom for R&D, so it will be interesting to see where they went with the final product.
 
They are that cheap and them some. That is why people grumble in these spec threads when Nintendo is concerned. They know their history and their stubbornness to change!
The reason people 'grumble in these spec threads' is due to individuals driving by, posting nonsensical 'lol, nintendo'..
 
Going from 720p to 1080p requires 2.25x the GPU power for rendering alone, so a game running on the XB1 @ 1080p may roughly run the same game at 720p with a GPU close to 600 GFLOPS (the XB1 is 1.31 TFLOPS. There is also the architecture differences, CPU, and memory setup to consider, but the 2x fp16 for Maxwell/Pascal's architecture and the rumored stronger CPU would give the Switch a performance boost beyond the documented GFLOPS for fp32 in comparisons to the other systems.

Having said that, I don't know if most multiplatform games ran @ 1080p on the XB1.

So, if Switch can do the 512 Gf at portable mode, that means it can do visuals very close to what a xbox one can do.

When docked, if 1 Tf is out of question, maybe a modern upscale to TV give the trick to made Switch visuals ok.
 
So if Nintendo did end up going with 20nm Maxwell for this version of Switch, could they later go with a die shrunk version and retain full compatibility while increasing battery life?
 
So if Nintendo did end up going with 20nm Maxwell for this version of Switch, could they later go with a die shrunk version and retain full compatibility while increasing battery life?
I don't see why not. 16nm Pascal for an iteration for 5-8 hours battery life(or increased performance) sounds plausible.. Assuming if Nate is wrong.
 
So, if Switch can do the 512 Gf at portable mode, that means it can do visuals very close to what a xbox one can do.

When docked, if 1 Tf is out of question, maybe a modern upscale to TV give the trick to made Switch visuals ok.
This is assuming the battery can last for 3(or 5-8 hrs from Nate Drake's sources) hours at 512TFLOPs in portable at something like around a 5 watt power draw.
I know its been mentioned a few times here in this thread and the venture beat article that it wouldn't need to use a fan if it's below 400TFLOPs(not sure where they got their example from.. Nvidia shield?), but Nintendo could go below that in portable to save battery.

Who knows.. Would like someone with legit tech knowledge to weigh on this.
 
...

PascalEdDay_FINAL_NDA_1463156837-008.png

Switch first real world application of middle out compression confirmed. Pied Piper saved.
 
So, if Switch can do the 512 Gf at portable mode, that means it can do visuals very close to what a xbox one can do.

When docked, if 1 Tf is out of question, maybe a modern upscale to TV give the trick to made Switch visuals ok.

Apparently the switch actually renders the image to a higher res(assumingly 1080p) when docked.
which is insanly awesome
 
They are that cheap and them some. That is why people grumble in these spec threads when Nintendo is concerned. They know their history and their stubbornness to change!

they're currently in one of the most substantial states of flux in their entire history, from software to hardware to merging their three main corporate wings under common goals. please try harder
 
Apparently the switch actually renders the image to a higher res(assumingly 1080p) when docked.
which is insanly awesome

Perhaps it would be right to assume docked mode is 2.25x more powerful than handheld since 1080p requires that much more power than 720p?

300 mobile, 700 docked(if we're lucky).

Given that Wii U's GPU is 176 GFLOPS and Nvidia's architecture is more moden, it will be easily 2x as powerful in mobile. So we should be seeing zelda u with stable framerates and extra effects.

People have also mentioned down sampling from dock to portable mode, but the dock mode increasing performance is something that is universallly agreed upon.

Anyway, what's practically guaranteed are:

-16-20nm node. Customized maxwell.
-at least 4GB RAM
-CPU not underpowered and won't be the bottleneck. More powerful than xbone and ps4 relative to its gpu when compared to theirs
-at least 1-2x wii u in portable mode, 3-4x wii u docked(or 512 GFLOPS or 1/2 xbone after counting newer and architectural advantages over amd)
-at least 3 hours of battery life
 
I want to be realistic about the power between both modes as I'm seeing in previous replies, but I also want to believe that Nvidia has some crazy black magic going on with this chip. The official blog literally says it's a custom chip with half a millennium's worth of time put into the engineering for it. I know this might be a dumb question, but do the developers supporting the Switch have used a dev kit before deciding whether to say they're supporting or not? I was shocked seeing FromSoft is on board, especially since they basically laughed when they were asked about porting Dark Souls 2 to the WiiU, even though the WiiU was plenty powerful though. (I think it was really because they believed the audience was along the younger audience, which seems opposite of what the marketing is going for with the Switch)
 
I want to be realistic about the power between both modes as I'm seeing in previous replies, but I also want to believe that Nvidia has some crazy black magic going on with this chip. The official blog literally says it's a custom chip with half a millennium's worth of time put into the engineering for it. I know this might be a dumb question, but do the developers supporting the Switch have used a dev kit before deciding whether to say they're supporting or not? I was shocked seeing FromSoft is on board, especially since they basically laughed when they were asked about porting Dark Souls 2 to the WiiU.

There is no black magic, but instead a well-thought-out form factor and marketability going on. And yeah, I think it's very safe to say that FromSoft had a dev kit before we even had serious rumors about it.

The Switch is not going to do well because of its guts. It's going to do well because of great design and versatility, things that matter much more to the mass market than do TFLOP numbers.
 
The Switch is not going to do well because of its guts. It's going to do well because of great design and versatility, things that matter much more to the mass market than do TFLOP numbers.
I think that's the key point here: support.

I hope the January showcase will announce a TON of games. This is the stretch for them that's literally do-or-die. Now that they've consolidated all their development groups (from 3DS and Wii U) to one platform, there are literally no excuses they should mess this up.

From the sound of things, the fact UE4 and other popular engines are on board seem promising. Will GameMaker also make the jump as well, I wonder? Is the fact that Nvidia is backing the hardware make it look more promising than its AMD console counterparts? Clearly this time power isn't going to be the big issue here.
 
So if Nintendo did end up going with 20nm Maxwell for this version of Switch, could they later go with a die shrunk version and retain full compatibility while increasing battery life?

yes, this is already very common in the industry.

Switch will most likely use maxwell, however, like its always been stated, its custom. Nothing says it can take stuff from the Tegra P1, pascal. A soc die is not very customisable by default. It's a premade die, not like a mother board. It has architecture and that isn't just plug and play tech.

SO If its already custom, as they stated. For a SoC that can be very large changes. Not similarly to traditional PCs. I'm thinking this thing will play between Tegra generations. Nintendo is pushing for this thing, sofor them to have waited even longer for Tegra Pascal back in the R&D phase a few years ago would have delayed this probably few more quarters if not a year. With this being nintendo, that extra power was not a concern that big.
 
I want to be realistic about the power between both modes as I'm seeing in previous replies, but I also want to believe that Nvidia has some crazy black magic going on with this chip. The official blog literally says it's a custom chip with half a millennium's worth of time put into the engineering for it. I know this might be a dumb question, but do the developers supporting the Switch have used a dev kit before deciding whether to say they're supporting or not? I was shocked seeing FromSoft is on board, especially since they basically laughed when they were asked about porting Dark Souls 2 to the WiiU, even though the WiiU was plenty powerful though. (I think it was really because they believed the audience was along the younger audience, which seems opposite of what the marketing is going for with the Switch)

Here's what I'm thinking in regards to ports

720p Switch, 900p Xbone, 1080p PS4

Would be interesting if Switch's CPU really was better than the other two. RAM and bandwith worry me though. CPU shouldn't be a bottleneck this time.

There is no black magic, but instead a well-thought-out form factor and marketability going on. And yeah, I think it's very safe to say that FromSoft had a dev kit before we even had serious rumors about it.

The Switch is not going to do well because of its guts. It's going to do well because of great design and versatility, things that matter much more to the mass market than do TFLOP numbers.

Agreed. As long as the switch flies off shells and it makes money for third party devs, third party devs will keep making games for the switch.
 
About resolution:

- 240p: 352 x 240 = 84.480 pixels (3DS resolution with 3D active- 800x240p with 3D off)
- 540p: 960x540 = 518400 pixels ( 5x the number of pixels of upper screen of 3DS)
- 720p: 1280 x 720 = 921.600 pixels (1,7x the number of pixels of 540p)
- 900p: 1600 x 900 = 1.440.000 pixels (1,5x the pixels of 720 or 2,7x the 540p)
- 1080p: 1920 x 1080 = 2.073.000 pixels (2,2 x the pixels of 720p or 4x the 540p)

I believe that is better Nintendo do a stronger portable and aiming for 900p on console.
 
About the battery (5-8 hrs) rumor, considering this is Nintendo we're talking about and their experience with handheld devices, isn't it possible they could have downgraded the chip to achieve this? It could be some variant of Maxwell 20 nm that is less powerful to achieve that kind of battery life.

Not saying this is is what I want. I'm hoping for 400-500 GFLOPS portable, 800 docked, but would be fine with 250-400.
 
About the battery (5-8 hrs) rumor, considering this is Nintendo we're talking about and their experience with handheld devices, isn't it possible they could have downgraded the chip to achieve this? It could be some variant of Maxwell 20 nm that is less powerful to achieve that kind of battery life.

Not saying this is is what I want. I'm hoping for 400-500 GFLOPS portable, 800 docked, but would be fine with 250-400.

The problem is according to rumors and the patent, there is a fan that runs in portable mode. That wouldn't be necessary if they downclock the chip so much so that the battery lasts 5-8 hours. I'm willing to bet that this 5-8 hour rumor won't pan out, and if it does then the SoC has to be on a 16nm process, regardless of architecture.
 
About resolution:

- 240p: 352 x 240 = 84.480 pixels (3DS resolution with 3D active- 800x240p with 3D off)
- 540p: 960x540 = 518400 pixels ( 5x the number of pixels of upper screen of 3DS)
- 720p: 1280 x 720 = 921.600 pixels (1,7x the number of pixels of 540p)
- 900p: 1600 x 900 = 1.440.000 pixels (1,5x the pixels of 720 or 2,7x the 540p)
- 1080p: 1920 x 1080 = 2.073.000 pixels (2,2 x the pixels of 720p or 4x the 540p)

I believe that is better Nintendo do a stronger portable and aiming for 900p on console.

I think it makes sense. I'd be surprised to see a 2.25x power increase when docked.
 
I want to be realistic about the power between both modes as I'm seeing in previous replies, but I also want to believe that Nvidia has some crazy black magic going on with this chip. The official blog literally says it's a custom chip with half a millennium's worth of time put into the engineering for it. I know this might be a dumb question, but do the developers supporting the Switch have used a dev kit before deciding whether to say they're supporting or not? I was shocked seeing FromSoft is on board, especially since they basically laughed when they were asked about porting Dark Souls 2 to the WiiU, even though the WiiU was plenty powerful though. (I think it was really because they believed the audience was along the younger audience, which seems opposite of what the marketing is going for with the Switch)


Not being a buzzkill, just tbh 500 man years isn't huge for a chip...That's 250 people working two years. The NVN API is likely included in that too.

As ground up designs can take hundreds of people four or five years, it sounds like it's definitely a modified existing design [which we knew, but to say, not dramatically different than what exists]
 
About the battery (5-8 hrs) rumor, considering this is Nintendo we're talking about and their experience with handheld devices, isn't it possible they could have downgraded the chip to achieve this? It could be some variant of Maxwell 20 nm that is less powerful to achieve that kind of battery life.

Not saying this is is what I want. I'm hoping for 400-500 GFLOPS portable, 800 docked, but would be fine with 250-400.

Yeah, that would be another way of looking at it.
 
There is no black magic, but instead a well-thought-out form factor and marketability going on. And yeah, I think it's very safe to say that FromSoft had a dev kit before we even had serious rumors about it.

The Switch is not going to do well because of its guts. It's going to do well because of great design and versatility, things that matter much more to the mass market than do TFLOP numbers.

Well, yea, but the guts are also important. They can be weaker than Xbox One, sure, but the system needs to be able to play the games. I'm just afraid of devs cancelling Switch ports because of problems getting the game to run well, which should be MUCH less likely now that they're using modern hardware, especially Nvidia hardware, compared to every Nintendo home console using a PPC CPU from the 90's since the Gamecube. I don't mind games rendering at 720p or sub (Halo 3 ran at 640p) when portable, especially if the Switch screen is 720p, and range from 720p to 1080p on TV. I don't disagree that the versatility will sell the system, but there's still going to be a cesspool that will say the Switch "sucks" just because it can't render at an equal IQ to PS4 or run as well.

If the marketing team is now focusing on the teens and young adults, and they're advertising it right off the bat as a home console where you can play the games anywhere like a portable system, then people are going to expect to play the games they would normally see on Xbox and PS4 wherever they want without hauling a console and TV with them.
 
I think that's the key point here: support.

I hope the January showcase will announce a TON of games. This is the stretch for them that's literally do-or-die. Now that they've consolidated all their development groups (from 3DS and Wii U) to one platform, there are literally no excuses they should mess this up.

From the sound of things, the fact UE4 and other popular engines are on board seem promising. Will GameMaker also make the jump as well, I wonder? Is the fact that Nvidia is backing the hardware make it look more promising than its AMD console counterparts? Clearly this time power isn't going to be the big issue here.

It's always going to be an issue. Of course UE4 and other engines are going to support the console, but it's not like the features it offers are going to magically become less expensive.

Unless the Switch is able to pull a Wii and sell like hotcakes because it offers something nobody else has. Which definitely isn't the case here because tablets already exist. This console will likely be weaker than the Xbox One even while docked, and it's releasing on the eve of upgraded consoles with 4K features. Nintendo plays the same song and dance every console release by claiming all the 3rd-party developers you've been hoping for are going to develop for them, and then nothing significant happens because the console is just too weak and they don't want to waste the time and energy making it run well when they have 3 other platforms to develop for. Just like with every Nintendo console, i'm expecting the vast majority of its sales to be from whatever 1st-party titles they can tease or promise at launch. Smash bros, 3D mario, the party games (Kart, Party), Splatoon, Zelda, ect.


If they pull a decent looking Metroid effort out of their asses, then i'll probably jump on board anyway. But this looks like the same cycle the Wii and WiiU went through playing itself all over again.
 
It's always going to be an issue. Of course UE4 and other engines are going to support the console, but it's not like the features it offers are going to magically become less expensive.

Unless the Switch is able to pull a Wii and sell like hotcakes because it offers something nobody else has. Which definitely isn't the case here because tablets already exist. This console will likely be weaker than the Xbox One even while docked, and it's releasing on the eve of upgraded consoles with 4K features. Nintendo plays the same song and dance every console release by claiming all the 3rd-party developers you've been hoping for are going to develop for them, and then nothing significant happens because the console is just too weak and they don't want to waste the time and energy making it run well when they have 3 other platforms to develop for. Just like with every Nintendo console, i'm expecting the vast majority of its sales to be from whatever 1st-party titles they can tease or promise at launch. Smash bros, 3D mario, the party games (Kart, Party), Splatoon, Zelda, ect.


If they pull a decent looking Metroid effort out of their asses, then i'll probably jump on board anyway. But this looks like the same cycle the Wii and WiiU went through playing itself all over again.

Its not primarily about power. The strongest system doesn't get the most amount of sales per generation. See PS1, PS2, Wii, all the nintendo handhelds vs Sony's.

It's about sales. As long as the system is hot and they get the sales to make profit, they will make games for Switch.

We've had call of duty games last generation on the Wii--a console that's like 20x weaker than 360/PS3, and the core gameplay was very much identical to its ps3/360/pc versions, with 30fps, no kill cams, and no controlled air kill sreaks.

If you look at the Wii and Wii U, not many 3rd party games sold a million or over. They didn't sell nearly as much as nintendo's 1st party games.

And Nintendo at this point is in a losing war if they did a power console with nothing radically new to compete against sony and microsoft.

Other than that 1/2 of xbone is not a big deal. Lowering the resolution and texture if need be, will happen.
 
Its not primarily about power. The strongest system doesn't get the most amount of sales per generation. See PS1, PS2, Wii, all the nintendo handhelds vs Sony's.

Completely different situations though. PS1 and 2 were in a time where the dominant guy (Nintendo) was an arsehole and Sony offered developers the chance to do things they wanted to do. That's not the case now, as they can do all they want on PS4/XBO.

Wii was popular with the mainstream and didn't really get any of the big third party support.

The handhelds is different, Sony just couldn't get into that space and those people that want "power" were on the home console anyway so handhelds didn't matter.

Switch is in another Wii/Wii U situations. Going for a mainstream audience that isn't the usual console crowd (Because Nintendo know they can't win that) and will primarily be Nintendo (And now handheld) games. The big AAA houses might try but really what's the point in getting a few more sales when you can put that time into making more money on the well established incumbents?

There is nothing about the Switch so far that should make someone feel this is going to be a AAA third home.
 
There is nothing about the Switch so far that should make someone feel this is going to be a AAA third home.

I would say the potential for truly portable console like experiences is a potential draw.

However, and it's a HUGE however, Nintendo thus far has not been a system for sports or FPS titles which are the types of AAA games that sell year over year and help establish customer bases for those companies.
 
Completely different situations though. PS1 and 2 were in a time where the dominant guy (Nintendo) was an arsehole and Sony offered developers the chance to do things they wanted to do. That's not the case now, as they can do all they want on PS4/XBO.

Wii was popular with the mainstream and didn't really get any of the big third party support.

The handhelds is different, Sony just couldn't get into that space and those people that want "power" were on the home console anyway so handhelds didn't matter.

Switch is in another Wii/Wii U situations. Going for a mainstream audience that isn't the usual console crowd (Because Nintendo know they can't win that) and will primarily be Nintendo (And now handheld) games. The big AAA houses might try but really what's the point in getting a few more sales when you can put that time into making more money on the well established incumbents?

There is nothing about the Switch so far that should make someone feel this is going to be a AAA third home.
The Switch could actually set up a "Play AAA games anytime and anywhere" audience to help make porting more ambitious games to the system a feasible investment. I don't think that is a farfetch idea if Nintendo plays things right.
 
The Switch could actually set up a "Play AAA games anytime and anywhere" audience to help make porting more ambitious games to the system a feasible investment. I don't think that is a farfetch idea if Nintendo plays things right.

Anything's possible but it doesn't seem like people prefer playing AAA games anywhere as opposed to playing with the best graphics they can get.
 
I have been all over spreading this need. Rocket league with custom Nintendo ip rocket trails and toppers would be amazing. The concept of the damn switch is perfect for this game. Local multiplayer dueling and co-op online with friends. No other platform is going to offer rocket league what the switch can do. Essentially rocket leagues window to mobile.

Imagine playing online at home then deciding to go over to a friends to match make online.
 
Anything's possible but it doesn't seem like people prefer playing AAA games anywhere as opposed to playing with the best graphics they can get.
I don't recall anything like what the Switch is trying to do, though. PSVita could have possibly evolved to that, but it didn't get enough western or first party support to become a real "portable console."
 
Status
Not open for further replies.
Top Bottom