StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (New Staff Post, Please read)

Absolutely, no "except" about it, though they figured out latency with Wii U, battery life and/or performance would suffer in reverse Wii U mode. I think it's unlikely but possible, if it does exist it's an entirely optional thing.
It's certainly possible it's just that I don't think it's as exciting an idea as people are making it out to be. And to the extent that it could work with a dongle it could also work with an iOS/Android/eShop app in the other direction with a docked Switch 2

The supposed second USB-C port is seriously the most overthought detail right now. If there is one, it's so you can charge the console in tabletop mode. That's it. There could be other benefits like using USB-C headphones, but that would be the reason for its inclusion. The fact that the Switch can't be changed in tabletop mode is a notable rough spot in the hardware proposition (that multiple third-party stands, and one first-party, were created to solve). Nintendo could have easily decided to smooth it out.
Worth noting that there's no shortage of Windows and Android based portables that do pretty much exactly this. One Type-C port wired for display output or connecting an external GPU and then a second which can just charge the device and is more conveniently placed

With that said, having a high speed USB port available to use while you're docked? Maybe that opens up some options. Who knows. Although given the Switch's USB ports were rarely used outside of third party controller dongles and LAN adapters I'm not holding my breath. Not much it opens up that can't already be done on the Switch as it is really
 
Just read the last threadmarked post.

Funny how we already (before confirmation ofc) know tons of details and characteristics of the system and are just kinda waiting for the official reveal.
 
While the bigger joy-cons will be better when used in a pair, I do wonder if the fact that the joysticks + buttons are still all on the same side won't make it harder to use single joy-con, especially for kids.

Of course I prefer the pair experience being prioritized, but this might be a negative side effect of making the joy-cons bigger.
 
0
Please give more meaningful contributions to this thread. Random quips like this are better suited for the Switch 2 Speculation ST. - meatbag
The new game console is just a legend.šŸ˜­
 
Cost matters, but making it appealing does to.
Of course - but that appeal for the Lite, like Nintendo 2DS, is entry price, not features. Adding extra features to Nintendo 2DS, like NFC functionality, required the purchase of a not-too-cheap external IR reader. This continued well into the lifespan of New Nintendo 3DS when the other non-NFC capable 3DS Family systems were discontinued. I hope that example makes it a bit clearer.
 
TSMC 7nm and Samsung 8nm may sound like they're "close", but they're really not.

Samsung 8nm is really more like a dressed up 10nm process.

The power draw would be substantially better on TSMC 7nm and the chip would be considerably smaller.
Yeah, I'm aware that SEC8N is actually optimized 10nm, like how TSMC 4N is optimized 5nm.

I just don't think there's enough of a reduction for TSMC 7N for it to make sense, but like I said it's not my area of expertise, I don't know how to do the math on that one.
 
And why the new Lite can't be about cheap handheld and TV mode this time?

The new Lite could be like a Series S which doesn't requires dev effort since the profile is already there.

The hybrid still has 4K, tabletop, bigger screen, better battery life, 2 controllers out of the box and other advantages.

The first Lite was the least popular model by far and it was 2/3 of the hybrid price. If it's 3/4 this time, then it will be even less popular if the proposition is the same.

A switch lite with tv mode would be so confusing .
 
A switch lite with tv mode would be so confusing .
I agree it would be confusing if it were including, but like how a Digital Edition PS5 (slim) can have a drive attached to it, it could make sense as a seperate accessory.

I think there's reasons for Nintendo to explore this idea this time, since the hybrid unit is so much larger, allowing people to buy all the bits to make a smaller future revision into a hybrid system might be very appealing for those with small hands, or children, or those who want to put their games on the TV without the need to buy a whole new system.
 
0
Sorry, not to pepper you with questions, but- what details overlap between the two leaks? When was the driver leak? I'm curious which details have been reconfirmed most recently and which details are still multiple years old and so have had more time to have possibly changed.
The driver leak was about 7 months after the NVN hack, but because the NVN hack doesn't come with dates from the code repository, and Nvidia clearly attempted to delete the Linux driver (and didn't quite cover their tracks), we can't be sure of the exact dates of either. Though we can be pretty sure both were code from pretty near the time of the leak.

What's interesting about the Linux driver is it includes the commit messages for the source code, several of which strongly indicate tape out of the chip happened sometime before August of 2022. The NVN hack happened in very late February of 2022. So there isn't a huge window for changes.

The Linux code doesn't include a GPU driver, just the driver for the rest of the chip. It doesn't include the exact size of the GPU, or it's architecture. However, there was a third, and a fourth leak from Nvidia. Twice they released documentation on their tegra chips that accidentally included T239. This confirms the GPU architecture as Ampere.

Other than the size of the GPU, the Linux driver confirms every big finding from the NVN hack. Both the NVN hack and the Linux driver include references to clock gating features backported from Lovelace. Both the NVN hack and the Linux driver show the 128-bit memory bus. Both the NVN hack and the Linux driver reference the number of PCIe lanes and the DisplayPort controller.

The NVN hack remains the sole source of the CUDA core numbers. The Linux driver remains the sole source of the CPU architecture. Neither of these things have been reconfirmed since, and for the most part Nvidia got their shit together, and locked down their accidental leaks in 2022.

There has been one other source of info about the chip since, however. A number of contractors for both Nvidia and Samsung have mentioned the chip in their resumes. They--

Wait, Samsung?!?! 8nm confirmed?!?!

No, wait a second. The Samsung contractors were not at Samsung's foundry services, but their custom IC group. These contractors worked on the cartridge technology, not the SOC directly. That's one of the sources for our knowledge about the new cartridges.

The Nvidia contractors have been electrical engineers. They've added no new details about the chip, but seem to confirm the tapeout timeline of Summer 2022. These resumes have circumstantial evidence for the process node - one of the engineers working on electrical verification also worked on RTX 40, which is on 4N. This is suggestive, but hardly a smoking gun.
 
We have an idea how it is built, really, from practical realities, safety concerns and Nintendo's past handhelds. It has to be designed with "user replaceable" batteries by law but that is a very low bar, as I've been discussing.

A physically large, less dense battery does not make any sense. They'd be paying more for the materials and getting nothing out of it, while giving people who want a bigger battery a not-so-simple task if they want to pay extra for a better battery. It would also be bad PR, like Wii U, with people wondering why they didn't just include the battery to begin with.
yes, a large less dense battery doesn't make sense. I've been on that train since the original post. hence the whole point of humoring a sellable larger capacity battery

and we still don't completely know how Drake is built. practices change over time and based on the hardware therein. we see no heat plate, no cooling, etc, so we can't make reasonable guesses as to how accessible the battery is until the final unit is shown and torn down
 
Took a screengrab from the 3d printed model video and cleaned it up a bit and turned the screens on to give a better representation of each experience.

Switch2-3dprint-cleanup.jpg


Edit: So, perhaps ignore the above in regards to the screen size. The above would represent a much bigger than 8" sized Switch 2 screen. I created this looking at the 3d print which includes ridges for what I assume would be the active screen area. Either the whole thing is printed out bigger then it should be, or those ridges are wrong and the active screen is smaller and therefore the device has much larger bezels.


Okay... so if you grant that the 3d printed mock up is correctly sized, this would be the corrected screen sizes/bezels for a 7" Switch and 8" Switch 2.

Switch2-3dprint-cleanup2.jpg
 
Last edited:
The supposed second USB-C port is seriously the most overthought detail right now. If there is one, it's so you can charge the console in tabletop mode. That's it. There could be other benefits like using USB-C headphones, but that would be the reason for its inclusion.

The fact that the Switch can't be changed in tabletop mode is a notable rough spot in the hardware proposition (that multiple third-party stands, and one first-party, were created to solve). Nintendo could have easily decided to smooth it out.
Spot on. It's either this or just there for development reasons (debugging for devkits or prototypes).

A DS style second screen accessory is not happening, as much as I would love it.
 
TSMC 7nm and Samsung 8nm may sound like they're "close", but they're really not.

Samsung 8nm is really more like a dressed up 10nm process.

The power draw would be substantially better on TSMC 7nm and the chip would be considerably smaller.

It is so much better that by comparison, its a far more feasible node for T239. I believe its around 35-40% more power efficient than 8nm. The main argument against it is that Nvidia doesn't really use it elsewhere. Its enough for the lowest clock speeds to actually work. We would be looking at Erista Switch levels of battery life, but at least you can make the math work, even if just barely. Samsung also has their 5nm process, but I believe that is just barely better than TSMC 7nm in terms of power efficiency and the yields were always said to be terrible. The node that makes the most sense for T239 is 4N, in terms of power consumption it fits the bill nicely and Nvidia already has plenty of capacity purchased with this node. That doesn't mean they did secure a deal for a different node specifically for T239, its possible, but from what I have read here, its generally more risky when you don't know for sure what demand will look like in the long term.
 
The top charging USB port on my Steam Deck is nice. Even in handheld mode, it is a nice touch to just have it not underneath you as you play.
 
It's certainly possible it's just that I don't think it's as exciting an idea as people are making it out to be. And to the extent that it could work with a dongle it could also work with an iOS/Android/eShop app in the other direction with a docked Switch 2
I agree! I think a mobile solution would probably be more practical.
 
0
The driver leak was about 7 months after the NVN hack
Thanks for the summary. I had been able to quickly find that the main leak was spring 2022, but I thought I remembered from (probably this thread) sometime in the last few months that the information in the leak was already pretty old. And with the leak being 2.5 years old at this point, I wondered if some of the details were more like 4+ years old without more recent reconfirmation. It doesn't sound like that's obviously the case, but we can't be sure.
 
There has been one other source of info about the chip since, however. A number of contractors for both Nvidia and Samsung have mentioned the chip in their resumes. They--

Wait, Samsung?!?! 8nm confirmed?!?!

No, wait a second. The Samsung contractors were not at Samsung's foundry services, but their custom IC group. These contractors worked on the cartridge technology, not the SOC directly. That's one of the sources for our knowledge about the new cartridges.

The Nvidia contractors have been electrical engineers. They've added no new details about the chip, but seem to confirm the tapeout timeline of Summer 2022. These resumes have circumstantial evidence for the process node - one of the engineers working on electrical verification also worked on RTX 40, which is on 4N. This is suggestive, but hardly a smoking gun.
I'm fairly certain no Samsung Linkedin profiles have mentioned T239. There are some that mention Nintendo projects which sound game card-related.
 
Took a screengrab from the 3d printed model video and cleaned it up a bit and turned the screens on to give a better representation of each experience.

Switch2-3dprint-cleanup.jpg
BTW, I used the 3d print as the guide for this screen size on the Switch 2 here. It appears if the OLED is a 7" screen this mockup would represent something a little bigger than an 8" screen, so... The bezels should be 1 or 2 mm bigger on the real deal. Small point but thought I would point that out. Perhaps the 3d print overall is a bit bigger then it should be? Could be a mistake or just an effect of a consumer grade 3d printer? Not sure. It does appear this guy is ever so slightly bigger then our discovered measurements.
 
The C (connect) button will clearly be used to activate the streetpass by carrying only the right joy-con šŸ¤¤

or...
Hidden content is only available for registered users. Sharing it outside of Famiboards is subject to moderation.
 
Spot on. It's either this or just there for development reasons (debugging for devkits or prototypes).

A DS style second screen accessory is not happening, as much as I would love it.
Maybe devs can conect a 2nd screen to debug the reverse wii u.
 
0
I'm fairly certain no Samsung Linkedin profiles have mentioned T239. There are some that mention Nintendo projects which sound game card-related.
You're right, I conflated Nintendo references generally with the chip itself.
 
You're right, I conflated Nintendo references generally with the chip itself.

Hidden content is only available for registered users. Sharing it outside of Famiboards is subject to moderation.
 
Last edited:
I just hold zero belief that theyā€™re gonna bring back that function. And I hold even less belief that the C button is for that purpose. Iā€™m getting to the point where I think itā€™s just a regular olā€™ button to be used in game. Cause it being a sleep mode button doesnā€™t make much sense.
Buttons are group by "intent" the button next to a system button is a system button . It has a letter C ... there are not many options what it can be... camera , cast, connect ...

also wasnt there a magnetometer or something like that in the switch 2? the same that was in the wii u and missing in the switch
I do not understand why people try to make it work so much as it only HINDERS gameplay in my experience. If I am playing a game, I don't want to move my head constantly to check the minimap or anything else. And asynchronous multiplayer is such a niche usage that they will not spend a dime to make it work imo.
The ir camera is very niche too , and now we have 2

so far the clearest and most convincing counterargument I've seen against it having some sort of casting capabilities is the rendering cost one (especially if it's trying to hit 4K on a second screen from handheld mode and not just, like, emulating DS games) and, sincerely, all the rest of them seem like... quibbles or complete non sequiturs. (an optional casting feature would depress third party support? no one wants to turn their head to look at a minimap?)
this mode would allow devs to split the handheld power between 2 screens... and if you use the top usb c charger while you play , full power.
 
Nintendo official report says that the Lite had similar margins as the hybrid, so they were profiting a bit less from the Lite, but the difference should still be close to $100.

And I will need a source on that "could have sold at profit for 249 in 2019". Keep in mind that the estimated $257 BOM doesn't means Nintendo pockets $42. There are other costs after that like shipping, warehouse fees, retailers cut, etc.


Sure, but that looks way less appealing, IMO.


I don't need to trust you. I'm not fighting over which one cheaper, you are. I'm saying they could include a stick with a Lite and continue selling hybrid docks for $60, regardless of how much it actually cost them.


And why the new Lite can't be about cheap handheld and TV mode this time?

The new Lite could be like a Series S which doesn't requires dev effort since the profile is already there.

The hybrid still has 4K, tabletop, bigger screen, better battery life, 2 controllers out of the box and other advantages.

The first Lite was the least popular model by far and it was 2/3 of the hybrid price. If it's 3/4 this time, then it will be even less popular if the proposition is the same.
It offered terrible VALUE. Because the regular switch offered $200+ more value from home console functionality but only costed $100 more. A switch tv only could have worked for $169 or less but people don't want a switch that is portable only when there's nothing stopping it from being dockable.
 
0
1. Trust me, USB C to HDMI adapters can be made much cheaper than that. Not that Nintendo would want it to be that cheap anyway.
Yeah, USB-C to HDMI adapters could be another solution to reducing costs, hypothetically speaking, especially since USB-C to HDMI adapters do have a DisplayPort to HDMI converter chip inside.
For Switch 1 they didn't even begin to entertain this idea anyway - they just made docking to a TV impossible.
I don't know if that's necessarily true. There was an article from Nikkei talking about how the Nintendo Switch Lite was (or planned to) still able to access TV mode. (I don't know how accurate Nikkei is in this case.)
 
I keep seeing that T239 is being used for automotive applications. Idk if that means it can't be used for the switch2 or not.
1. I suggest not linking to people's Linkedin profiles. It may be public, but it's still personal information and shouldn't be idly shared around.
2. Neither of those profiles say anything about T239 being used for automotive applications.
3. Even if T239 was used for automotive or other applications (which it isn't), Nintendo, the customer it was designed for, would still use it.
 
I keep seeing that T239 is being used for automotive applications. Idk if that means it can't be used for the switch2 or not.

[snip]

Please don't directly link to people's LinkedIn profiles. We don't want to encourage stalking or bring excessive attention to people who worked on things related to Switch 2, and may not have realised they weren't supposed to include it on their LinkedIn profile.

If there is anything relevant, you can post the relevant parts (in hide tags), and then if anyone needs to verify they can PM you for the link.
 
Speculation months ago from some of the guys at DF were suggesting that the increased display size was to help accommodate a large battery. Recent discoveries are certainly not pointing to a large battery. This totally undermines the basis for which that speculation was founded. Big screen, big battery, and big SOC. Big SOC with a little battery doesn't work, and a small battery with a big screen means the larger screen size has nothing to do with housing a large battery and SOC. I'm all ears for a compelling argument advocating for 8nm if you can offer convincing solutions to these problems.
 
1. I suggest not linking to people's Linkedin profiles. It may be public, but it's still personal information and shouldn't be idly shared around.
2. Neither of those profiles say anything about T239 being used for automotive applications.
3. Even if T239 was used for automotive or other applications (which it isn't), Nintendo, the customer it was designed for, would still use it.
Collaborated with cross-functional teams to design and automate regression and reboot test suites using python scripts across
128/256/512GB memory modules for T239/Grace for performance characterization for NVIDIAā€™s automotive customers.

I did see that atlan got canned or maybe repurposed to T239?
 
Collaborated with cross-functional teams to design and automate regression and reboot test suites using python scripts across
128/256/512GB memory modules for T239/Grace for performance characterization for NVIDIAā€™s automotive customers.

I did see that atlan got canned or maybe repurposed to T239?
I think the main thing I'd infer is that T239 and Grace share a memory controller.
 
Took a screengrab from the 3d printed model video and cleaned it up a bit and turned the screens on to give a better representation of each experience.

Switch2-3dprint-cleanup.jpg
Those joy con still donā€™t look quite right to me
Like they look wonky where as the og looks well designed

I know the 2 is just a mock up or printed or whatever ā€¦ but Iā€™m guessing the real deal will have better looking joycons somehow
 
Do we know if the console will have a unique gimmick to set it apart from the current Switch? I believe it needs something special, otherwise it will simply be a larger, more powerful Switch
No real way of knowing based on the current intel. Never put it past Nintendo.
 
C is for connect. Like connecting to the Internet or to your friends.

(I REALLY want a better online experience for the Switch 2. I really do.)
 
Do we know if the console will have a unique gimmick to set it apart from the current Switch? I believe it needs something special, otherwise it will simply be a larger, more powerful Switch
As of right now, no new gimmick is known which imo isnā€™t needed.
The PS4 was a more powerful PS3, and the PS3 was a more powerful PS2, etc. They all were successful.
A new gimmick isnā€™t needed because letā€™s be real, it would be a gimmick that gets quickly abandoned.
 
No real way of knowing based on the current intel. Never put it past Nintendo.
I get it... we'll have to wait and see. I'm concerned that people might look at it and assume nothing's different.

As of right now, no new gimmick is known which imo isnā€™t needed.
The PS4 was a more powerful PS3, and the PS3 was a more powerful PS2, etc. They all were successful.
A new gimmick isnā€™t needed because letā€™s be real, it would be a gimmick that gets quickly abandoned.
I agree, but you would lose the uniqueness of Nintendo that has always distinguished it.
 
Do we know if the console will have a unique gimmick to set it apart from the current Switch? I believe it needs something special, otherwise it will simply be a larger, more powerful Switch
Perhaps it is simply a larger, more powerful Switch.

The only gimmick that we have any real evidence for is... Maybe additional IR Motion Cameras? But that's pretty tenuous.

Being a "Switch but better in every way" I think is extremely appealing.

It helps that Nintendo effectively has a monopoly on handheld consoles at the moment. šŸ˜…
 
Please read this new, consolidated staff post before posting.

Furthermore, according to this follow-up post, all off-topic chat will be moderated.
Last edited by a moderator:


Back
Top Bottom