Switch is roughly the same size of the new smaller basic Shieldtv model but it is only half an inch thick versus the shieldTV being an inch thick.
-
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 Switch Dev Kit Stats Leaked? Cortex A57, 4GB RAM, 32GB Storage, Multi-Touch.
- Thread starter Robert at Zeboyd Games
- Start date
- Status
Switch is roughly the same size of the new smaller basic Shieldtv model but it is only half an inch thick versus the shieldTV being an inch thick.
lwilliams3
Member
Thanks. And the Switch has a screen and runs on batteries. Considering the little space they had, can we be sure that they didn't use 16nm to reach the base performance that we are looking at?
Well I agree your tests suggest a similar node and power draw and certainly its going to be based on that SoC. But I still expect it to be custom, more so than just input/output connectors. I think if it was simply a TX1 they'd have called it a TX1. Even if its just a few small changes within the GPU and a cache redesign, its going to be a custom SoC.
Just to clarify on size too. The Joycons certainly make it larger but all the hardware is within the tablet so feel its more comparable.
Basic Shield TV 2017
3.858 inches Height
6.26 inches Wide
1.02 Inches Depth
Estimated Tablet Size of Switch
4.19 inches Height
5.83 inches Wide
.55 inches Depth
Basic Shield TV 2017
3.858 inches Height
6.26 inches Wide
1.02 Inches Depth
Estimated Tablet Size of Switch
4.19 inches Height
5.83 inches Wide
.55 inches Depth
Shield TV 2017:-
Height: 98 mm (3.85 inches)
Width: 159 mm (6.259 inches)
Depth: 25 mm (1.01 inches)
Switch*
Height: 106 - 115 mm (4.17 - 4.53 inches)
Width - 184 - 196 mm (7.24 - 7.71 inches)
Depth - 15 mm (0.59 inches)
*Size based on estimates:
http://imgur.com/a/u6Vhp
http://arstechnica.co.uk/gaming/2016/10/how-big-is-the-nintendo-switch-screen/
EDIT Hadn't seen your post before posting Antonz, so not trying to correct you. Though I think the number you have for Switch width is a mistake in the arstechnica article.
Height: 98 mm (3.85 inches)
Width: 159 mm (6.259 inches)
Depth: 25 mm (1.01 inches)
Switch*
Height: 106 - 115 mm (4.17 - 4.53 inches)
Width - 184 - 196 mm (7.24 - 7.71 inches)
Depth - 15 mm (0.59 inches)
*Size based on estimates:
http://imgur.com/a/u6Vhp
http://arstechnica.co.uk/gaming/2016/10/how-big-is-the-nintendo-switch-screen/
EDIT Hadn't seen your post before posting Antonz, so not trying to correct you. Though I think the number you have for Switch width is a mistake in the arstechnica article.
I really hope they increased the memory bus width. If the Switch uses the standard 64-bit bus that Tegra X1 has, that means Nintendo has stuck with a 64-bit memory bus for 5 console generations in a row (Nintendo 64, GameCube, Wii, Wii U and now very likely the Switch, unless they took elements of Parker like the 128-bit bus).
Although this has absolutely no bearing on Switch (just a bit of trivia, no pun intended) Nvidia introduced a 128-bit memory bus in 1997 with the RIVA 128 card.
Its better to have a narrower bus at a higher frequency than a wider bus at a lower frequency (given the same overall bandwidth). So increased clock frequency is the better way to go compared to widening the memory bus.
Of course considering we already know Switch's memory speed obviously the wider the bus the better. Still think we'll get some kind of new cache design in the SoC that will alleviate main memory bandwidth though.
Of course considering we already know Switch's memory speed obviously the wider the bus the better. Still think we'll get some kind of new cache design in the SoC that will alleviate main memory bandwidth though.
I do actually agree with you looking at the Width. A quick look at 6.2" LCD has shown on average 155mm width. So yeah I would say at a minimum looking at 7 inch width because they are completely missing not only bezel but the extra space around the bezel
I wonder when Nintendo is going to add Switch to the Developer portal. It's still just Wii U and 3DS, no mention of the Switch anywhere.
Probably closer to launch or even after.
Memory bus width is based on the actual bus width of the RAM multiplied by number of parallel chips. It is not a spec that improves with technology as a wider bus also means more board traces and memory chips - something that does not increase with time. Bus width can be better correlated to cost than technical generation.
I wonder when Nintendo is going to add Switch to the Developer portal. It's still just Wii U and 3DS, no mention of the Switch anywhere.
Definitely not before the presentation since anyone can create a developer account and read all the dev documents, which would reveal all the specs right away.
LordOfChaos
Member
I have a feeling you're right about the lighter modifications, the self admitted 500 people years (250 people x 2 years) really only sounds impressive outside of practically any modern chip, if you include making a whole new API in there (even with Vulkan/Mantle as a base to work off), it's really not much. On a scale of TX1 to fully custom I'd expect it to be much closer to the TX1 side then.
If there's a lot more to it, I'd happily be wrong!
And agreed on memory too, it's been Nintendo's concern since N64 and it may be one of the things they touch on. This is a good thing, bandwidth is where mobile has been strangled even while the chips themselves superseded the last generation of consoles in power.
Maybe they meant 500 man years in terms of time charged? If that's the case, it's more like 8.5 years for 250 people. Usually man-hours and man-years are quoted in time worked, not time employed, though I could be wrong.
That sounds like a hell of a lot, I doubt it. I think a 250 person team working for 2 years sounds.a lot too
Im on team lightly modified TX1.
I'm sure its not going to be nearly as custom as WiiU's SoC. But I'm still going for significant modifications. I mean Pica200 had a whole set of instructions and eDRAM added and Nintendo still didn't consider it to be a custom GPU. I see no reason for them to claim this SoC is custom if that's not the case.
Only 12 hours left until the reveal I suppose. I mean I don't expect specs to be announced but surely there'll be a lot of developer interviews and we may find something out..
Only 12 hours left until the reveal I suppose. I mean I don't expect specs to be announced but surely there'll be a lot of developer interviews and we may find something out..
Mostly Nintendo mentioned modestly but in reality it is more than modest.
Shield TV is 26mm thick (actually 1mm thicker than Shield TV from 2015) Switch is 15mm thick, so I don't know where you got that info from but Switch has a much smaller area to cool a similar power consumption thanks to the need to charge a battery and the screen... which I didn't factor into the estimation of 13watts, so there is that...
PS the battery takes up space, it won't be available for airflow, and it's going to be something like twice as big as the 3DS batteries. I'm also still thinking the idea that Nintendo and Nvidia would order 10s of millions of switch chips at 20nm is pretty crazy, but if it can work, then that is what it is. I'm just skeptical that the smaller switch can cool as well as the larger shield tv.
I didn't say it was as thick, I'm talking about the overall size. Switch is slimmer, but its also taller and wider than the new Shield TV. Most estimates have it somewhere between 80-90% the overall area of Shield TV.
LordOfChaos
Member
Which would scale comparisons to any other chip making accordingly, in which case I wouldn't have to change my point. I don't think it's that though as it would scale chips that took 600 engineers 5 years to crazy proportions:
http://www.investopedia.com/terms/m/manyear.asp
Anyways, we don't really know if that's all engineers, or a mix of marketing and management, or if all of those 500 man years went to Switch or some were including progress of general GPU technology, etc.
Okay, I'm surprised that for how many times this has been mentioned nobody has said anything but... I don't think that 250 people were working 24/7 for two years straight. Also, everyone keeps comparing this to making a new GPU core from the ground up...
Edit: Never mind on the first part, but still does it really take 2 years to just decide on clock speeds? I don't think it does or else skipping 20nm would have been a much bigger setback for AMD and Nvidia.
The 3ds XL battery is 1800mah, I expect Switch's battery to have a larger capacity and be physically larger as well.
Height: 68mm
Width: 38mm
Depth: 6.7mm
This is a 3rd party battery because I couldn't find dimensions for the first party battery (they should be the same size, since they fit in a compartment on the device and have the same capacity) and am not at home to measure them myself. It's a lot of dead space inside the device that we aren't taking into account, the depth is already an issue, as you effectively cut air space in half, so yes if X1 is throttling, there is some hard questioning on whether the switch can actually cool the 20nm version of X1.
https://www.amazon.com/dp/B0121SAK5I/?tag=neogaf0e-20
LucidFlux
Member
You're right. If you look at Nvidia's blog post, it seems the 500 man-years covers the entirety of their involvement in switch's development, not solely the chip. Who knows how much of that 500 went towards modifying the SoC. I guess we'll see soon enough. Here is the quote from Nvidia:
"But creating a device so fun required some serious engineering. The development encompassed 500 man-years of effort across every facet of creating a new gaming platform: algorithms, computer architecture, system design, system software, APIs, game engines and peripherals. They all had to be rethought and redesigned for Nintendo to deliver the best experience for gamers, whether theyre in the living room or on the move."
LordOfChaos
Member
Creating a bespoke API isn't insubstantial.
Yeah there's no doubt that there'll be less space inside to some degree. I just think the fact Nvidia have made the new Shield so much smaller shows that internal space was never any issue in the original Shield TV as far as throttling goes. Which kind of makes sense when you look at the cooling solution they used. Will a bit less space again necessarily cause an issue?, I don't know with any certainty I suppose.
I mean I'm not saying I think the SoC is definitely 20nm, it could be 16nm who knows.
They also say it's a custom chip, the 500 man years is pointless to speculate on, because you could bulk all Tegra's development itself into it as well. It's just not a useful number.
No, Mdave figured out these Clock speeds are a good sweetspot for a tx1 in a few hours lol. They obviously did more than that.
Astral Dog
Member
We are going to see the result of 500 Man Years condensed into one in just a few hours 😆
I know, but the thing is that what you guys are talking about is an off-the-shelf X1 with lower clocks, and I don't think anybody would call that custom at all. How much work does it take to change the number of cores, really? I also highly doubt that the API is from the ground up, but if it is it would indeed take up pretty much all of that work, I admit.
And it still doesn't change the fact that it's silly to compare it to making a new GPU architecture.
Most expect an enchanced memory subsystem. That takes some work.
LordOfChaos
Member
A lot of the things you're saying we're talking about aren't what we're talking about. A new memory hierarchy (as mentioned) for instance wouldn't be insubstantial either. I'm just saying I think if TX1 was a 1 on an arbitrary scale and fully custom was a 10, I'd expect this to be a 3, kinda thing.
Comparing the work involved for a fully new GPU is a fair point just to say it's not going to be a substantially different uArch.
Yeah, but not too much. It would either be the addition of eSRAM or doubling the bus. It would technically count as custom though, I guess.
Well, nobody was expecting it to be fully custom, and anyone expecting a new architecture is, frankly, stupid. Do we know for sure that the PS4 and Xbone APUs took longer to develop? I wouldn't call those "fully custom" either btw.
For the record, I'm not really expecting more customization than that; I'm just pointing out that the man-hours thing is being downplayed too much and that comparing it to making a new uArch is unfair because nobody in their right mind expects anything like that.
The 3DS has a limit of 32GB, my 64GB works fine.
There is no hardware limitation for 128GB that I know of, it should work with a theoretical 2TB MicroSD card.
The amount of RAM per chip can increase down the road too. We're at 16Gbit density for LPDDR4. There exists 32Gbit for LPDDR3. Just a matter of time.
a 720p screen is very bad for VR. I had a nexus 4 with a 720p screen and google cardboard, although I liked the novelty, is not a very good solution because the screen is not that very dense for a good (note I didn't say perfect, just good-ok) immersion.
Also I think the screen with the bezel may be too big wearing it as a VR helmet comfortably
WiiU and 3DS could read bigger cards (and access all the storage space) than the ones that were "officially" supported. But we'll see...
calc84maniac
Member
Screen size is fine, don't forget dat bezel
Yeah I meant with the bezel. I'll edit it
prag16
Banned
Yeah. 720p would be terrible for that. I have a cardboard with an LG G4 (1440p) and the individual pixels are still noticeable even in that case.
Smurfman256
Member
1920x2160 per eye is probably going to be the resolution in which people stop complaining about the screen-dooring.
While it's true that the time invested is in reference to the entirety of their involvement with the Switch, the statement - if taken at face value - also suggests a redesign of the architecture itself. If we are to take those comments very literally, it all sounds well beyond simply tinkering with clocks. They clearly went out of their way to make this thing as efficient as possible, so I can't imagine them and Nintendo looking at that X1 and not seeing several ways to improve it. Heck, lowering the clocks may have well given them even more incentive to seek efficiency/gains in other areas of the chip.
I'm on team-moderately customized
(Supposing: Wii U gpu = highly customized, 3DS Pica gpu = lightly customized).
But yes, hopefully devs/engineers will be a bit more forthcoming with details in the coming days/weeks. Either way...hyped!
prag16
Banned
That should be "good enough" for most I'd think, but I'm sure some (nitpickers or the eagle eyed) will complain until it gets to 3840x4320.
LordOfChaos
Member
I tried VR on my iPhone 6S, 750P on 4.7", and it was godawful. Since resolution is haled per eye and it's so close to your face it was very much like sitting right in front of a 90s CRT SDTV.
That said, I think the Switch could be ok for cutesy VR demos, rather than anything you keep on for a while.
I'm not sure it's worth going to such lengths (particularly as it would be hard to infer too much from a GPU clock that's not perfectly stable).
I suspect the demos are triple-buffered anyway (no tearing on my end), but I'll check when I'm back on my PC. I probably should also note that they run windowed on PC, so when I say 1080p it's a maximised window so the actual resolution being rendered is a little less (maybe by 5-10%).
There's an iFixit teardown of the Pixel C, and if you go down to step 10 you can see the logic board, including the TX1. Comparing to the DRAM chips above it, it would seem that the 11mm x 11mm measurement is accurate, so 121mm² it is. (I had misremembered this as the measurement for the package, but that certainly looks like an exposed die).
I'd be a bit surprised if it's not IPS. Sourcing of 3DS screens was constrained by the 3D tech, whereas now it's a straightforward, standard resolution panel, so they shouldn't have any issues finding suppliers. The other thing is that TN screens seem to have pretty much disappeared at this kind of screen size, with even budget phones almost all using IPS screens these days. IPS is effectively the "cheap" phone display technology these days, with OLED replacing it as the more expensive option.
The one thing we really should be looking out for on the display front, though, is adaptive sync (i.e. G-sync, but probably without being called G-sync). For a fully integrated device like Switch, where Nintendo gets to choose and/or design everything from the OS to the APIs to the GPU to the display controller to the panel itself, supporting adaptive sync isn't actually all that expensive. There's up front R&D cost, to be sure, but Nvidia have obviously already done most of the R&D necessary. From that point it's a matter of using a display controller which can properly support adaptive-sync, and then getting all your software ducks in a row.
My only worry about this is that because Nintendo's internal software teams do an exceptionally good job of locking frame rates to either 30 or 60fps they might not see the value of adaptive sync. For third-party games it could be a huge benefit, though. Those inevitable not-quite-30fps ports would feel a hell of a lot smoother on an adaptive sync display than a fixed refresh display even if the actual frame-rate is identical. Compared to using a bigger GPU (which costs money) or using a higher portable clock speed (which costs battery life), adaptive sync is an exceptionally cheap way to give players a smoother gaming experience in portable mode, so let's hope they go with it.
I don't know how much of a bespoke API they've created, to be honest. In fact, compared to past hardware, where they actually have needed to create APIs from scratch (or nearly from scratch) to accommodate the new hardware, they could get themselves into a far better software and tools position than they've ever been in before with far less work involved.
From a graphics API perspective Vulkan already fits pretty much all their needs, and given they've joined the Vulkan working group and certified Switch as conformant to the Vulkan spec, it would certainly seem like they're using it as their main graphics API. Regarding tools and development, Nvidia already has a full (and reportedly excellent) set of development and debug tools which should be usable for Switch with minimal customisation. Then Gameworks includes plenty of graphics and physics libraries which should be usable on Switch with, again, relatively minimal work.
I suspect the 500 man-years probably does include quite a lot of software work (possibly even including work Nvidia was already doing anyway, such as bringing Gameworks to Vulkan), but if we're comparing the amount of work Nintendo and Nvidia need to put into software for Switch versus what Nintendo would have needed to do to get previous consoles to the same level, then Switch would be a bit of a cakewalk for them.
- Status