StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (Read the staff posts before commenting!)

[Instro]

So bets on when we first hear something official about this device?

Working backwards from something like a September/October-ish 2024 release, I'm penciling in next February. A peak behind the curtain so to speak. That gives them time for some directional stuff in their investor meetings, and then a big blowout in June.

The only way I see something happening this year is if "2nd half 2024" is more of a 3rd party target rather than a 1st party target, or it's very early on the 2nd half like June/July.

 

[entut1]

The Switch 2 launch title I want is that Wii space exploration game Nintendo showed once that never came out.

 

[Deleted member 887]

It's only listed as sampling, though, and supports speeds up to the full 6400MT/s, which should allow it to run a lot faster than on Orin Nano. The website could be out of date, and Nvidia could be limiting speeds for other reasons, like power consumption or product segmentation, or this could be a different module which shares the second part of the product code (I have no idea what 2UB47 refers to).

Nvidia is pushing the floor sweeping and binning of Orin pretty far. They've actually retroactively removed features from the spec sheet in order to increase yields. I would assume that binning extends to memory controller performance as well as the CPU and GPU, and that the downclock is a product of that.

 

[John Omaha]

It’ll be the new control gimmick of the Switch 2 like what HD rumble was for the Switch 1.

It's not a gimmick, though.

 

[Dekuman]

It's not a gimmick, though.

The gimmick will be the friends we made along the way

 

[chocolate_supra]

The gimmick will be the friend codes we made along the way

Fixed

 

[ILikeFeet]

The gimmick will be the friends we made along the way

pointless if Nintendo doesn't improve their online services and add online to games that only have local multiplayer

 

[Dakhil]

While technically Mariko was upgraded to support LPDDR4X, I don't know if there were any actual silicon changes to the memory controller to support this. LPDDR4X is simply an extension of the LPDDR4 spec with support for higher clocks and lower voltages, and the die shrink alone may have been enough to provide support without changing the silicon.

Patrick Moran mentioned that LPDDR4X isn't backwards compatible with LPDDR4 since LPDDR4X devices aren't required to be compatible with or tolerate higher LPDDR4 voltage on the I/O, and there's no specified way to query a LPDDR4 or LPDDR4X device to determine the type during the power-up sequence.

 

[Foltzie]

pointless if Nintendo doesn't improve their online services and add online to games that only have local multiplayer

Nintendo we want you to play with your friends, but no you can't help you make new friends.

 

[Foltzie]

Patrick Moran mentioned that LPDDR4X isn't backwards compatible with LPDDR4 since LPDDR4X devices aren't required to be compatible with or tolerate higher LPDDR4 voltage on the I/O, and there's no specified way to query a LPDDR4 or LPDDR4X device to determine the type during the power-up sequence.

Does the controller itself need to be updated to send the lower voltage or can that be handled in firmware.

This is not as bad as it sounds since LPDDR memory devices are not user (or even technician) replaceable once installed by the manufacturer.

 

[ILikeFeet]

Nintendo we want you to play with your friends, but no you can't help you make new friends.

discord groups are close enough!

 

[Quangcute03]

Well...Maximal Resolution may actually be limited by the ROP/Pixel Fill Rate rather than anything GPU related.

Assuming Thraktor's estimated clocks, and the sheer architectural advancement over GCN1.1/2 in the PS4/XBO, Switch 2 docked should be far closer to Series S/PS4 Pro GPU-wise...assuming ROPs/Memory/Pixel-Fill Rate don't become a problem.

Considering leaks are coming out indicating that Memory may actually not be a object with it maybe having more RAM than Series S even, and how GDDR6 and LPDDR5 can pretty much trade blows in an optimized setting due to the latters' far lower latency. ROPs/Pixel Fill Rate likely will be the bottleneck if anything at this point.

That's why I bring up the ROPs problem before. I feel 16 ROPs isn't gonna cut it.

 

[Starbreaker]

Knee IP? like a sequel to ARMS?

You got me lol.

 

[VegiHam]

You guys' spine had Lumas on it? I'm jealous. Euro one was a regular Wii spine

 

[Thraktor]

LTTP: I don’t know whether the Switch NG will introduce a new Game Card format. But if it does, I’m confident that the card will be backward compatible. The reason is simple: Nintendo will continue releasing cross-gen games for perhaps another 2-3 years. Unless they are going to make all cross-gen titles digital only (improbable), the Game Card of a cross-gen title will need to support both generations, therefore the new card format must be backward compatible when inserted into the older Switch models. Otherwise, Nintendo would need to release two SKUs for the same game; the partners and retailers certainly would balk at that, and some consumers probably would be confused by two packages of the same game too.

And if the new Game Card itself is backward compatible, the hardware costs of supporting the old card format on Switch NG should be fairly low too (same card slot, possibly with some extra pins). In more than one occasions, Furukawa has stressed the importance of migrating the Switch user base. It’d be an easy win to support the existing physical ownerships.

Your rationale holds up, I'm just not sure how they do it. Presumably replacing the NAND used inside the existing Game Cards doesn't require waiting for the new console. So if they're waiting for the new console, that implies a change to the existing physical bus, for speed. Usually the solution there is a second set of data pins, but the Game Card already has second row.

I'm on team "don't need an extra set of pins" at this point. Adding extra pins for physical interface changes often leads to monstrosities like USB 3 micro B and SD Express. Besides, those standards end up like that because of a desire for cross-compatibility across a wide range of products, but Nintendo only has very specific compatibility needs; they need Switch game cards to work in Switch NG, but they don't need Switch NG cards to be backwards compatible with the older interface, nor do they need to worry about compatibility with any third-party products.

Here's the Switch card pins, courtesy of Switch Brew:

There are 8 data pins (6,7,9,10,12-15), which is more than enough for any modern high-speed interface (it would allow up to about 4GB/s if they were used for PCIe 4.0 or about 3GB/s if using M-PHY 4.1). The question is how to use those pins for both the existing interface and the new interface and switch between the two. One option would be to start on the original SPI interface, and have a way of negotiating up to the faster interface on Switch NG cards, which I believe is how the newer USB and SD interfaces work. To me, though, this seems like it would just increase the complexity (and therefore cost) of the Switch NG cards, as they would need to support the old SPI interface as well as the new interface. My guess is that Nintendo will choose whatever approach minimises the cost of Switch NG cards.

I suspect it may be possible to cleanly handle a new card format with the addition of one new pair of pins. On the left of the pin diagram, you can see pin 1/2. This is a single physical pad on the cards, but two pins on the reader, and it's used to detect when a card is inserted; if the two pins on the reader are electrically connected by the card, then there's a card inserted. If the Switch NG cards are a bit wider (I'm assuming there will be some physical change in any case, to prevent them being inserted into the original Switch), then an extra pair of pins could be added next to the existing pins (let's say pins -1/0 or pins 18/19) to act as new card detect pins. Switch NG cards would have a pad connecting these two pins, but not pins 1 and 2, so the card reader would immediately know which interface to use when a card is inserted; if pins 1 and 2 are connected, it's the old interface, if pins -1 and 0 are connected, it's the new interface. New cards wouldn't need to deal with the old interface in any way, and therefore could be kept as simple (and cheap) as possible.

Another option is actually a pretty simple one; just have Switch and Switch NG cards insert the opposite way around from each other. Sony do this on their high-end cameras which support both CFexpress Type A and UHS-II SD cards. The two formats are completely incompatible, so they have a set of pins for CFexpress on one side of the reader, and a set of pins for SD on the other side, with CFexpress cards going in one way round and SD cards going in the other way. Nintendo could adopt a completely different pinout for the new game cards and just design the slot so that they get inserted the other way around from Switch game cards.

 

[Foltzie]

Wild, I never knew that so "few" consoles fit in a shipping container, at about 10-12.5K per 40ft container. Or about 75-100 containers per million consoles, or about 1500 -2000 pallets per million consoles.

An average 40ft container has ~70 cubic meters or ~2467 cubic feet. Assuming the Switch 2 box is the same size as the OLED box at 320 cubic inches (cu in) (0.18 cu ft) or the OG box at 412.5 cu in (0.24 cu ft), you could fix ~10-12.5K units per container. 20 pallets per 40 ft container.

Really highlights how storing stock becomes tricky from a logistics standpoint, let alone a cost standpoint.

Started fiddling with numbers for air freight vs containers and went down a rabbit hole.

 

[Samuspet]

So the Switch Lite was a bit of a letdown for Nintendo especially in Japan right ? which would have been it's main target market I imagine. I know this 99% won't happen but imagine if instead of offering a mobile only option for Switch 2 they offered a home console only version and here's the twist, both would be available at launch much like the Series S and Series X.

Just for fun could anyone do the math and see what the maximum clocks, performance and tflops Drake could output with say a 200w power budget instead of sub 20w?

I guess it could go -

Switch 2 Hybrid - 720p mobile / 900p docked then DLSS up to 1440p.
Switch 2 Home - 1080p then DLSS up to 4k. Much more performant Ray Tracing.

$399 for the Hybrid / $499 for the Home (Joy-Cons replaced with Pro Controller 2).

 

[ShadowFox08]

have a feeling this action RPG Monolith Soft is suposedly working since 2017, will be Xenoblade Chronicles 4 or a Xenoblade Chronicles X sequel, i dont see Nintendo aproving Monolith Soft to do a new IP, but Xenoblade Chronicles 4 or a Xenoblade Chronicles X sequel would be more intereting in Nintendo eyes

Xenoblade Warriors in 2024. Believe.

Jetson Orin Nano is an interesting one, as it the 8GB board claims to have a 128-bit LPDDR5 bus (with clearly two modules on there), but the speed is listed at 68GB/s, which is what you'd expect from LPDDR4X. The 4GB model has a 64-bit bus, again claiming to be LPDDR5, but with a speed of 34GB/s, which again is LPDDR4X performance. It only has one memory module on the board. So, the specs indicate they are both using 4GB 64-bit LPDDR5 modules, but the speeds suggest it's actually LPDDR4X.

Looking at the memory modules on Serve the Home's photo, they're made by Micron, and have the marking 2UB47 D8BFW. I haven't been able to find a part matching the full code, but this module is a 64-bit 4GB LPDDR5 part which lists the FBGA code as D8BFW. It's only listed as sampling, though, and supports speeds up to the full 6400MT/s, which should allow it to run a lot faster than on Orin Nano. The website could be out of date, and Nvidia could be limiting speeds for other reasons, like power consumption or product segmentation, or this could be a different module which shares the second part of the product code (I have no idea what 2UB47 refers to).

It could really just be lpddr5, but it's clocked so low it's at 68 GB/s so save power draw (4266.7 MT/S)

 

[SoldierDelta]

Xenoblade Warriors in 2024. Believe.

Xenoblade is the only series of Nintendo's that can fully justify a warriors cross-over game and I will not accept a counter-example.

 

[Deleted member 887]

Just for fun could anyone do the math and see what the maximum clocks, performance and tflops Drake could output with say a 200w power budget instead of sub 20w?

It would catch on fire. It’s simply too small to push 200 W through. And at 70 W you’re past the point of getting any returns anyway - basing this on some Ada overclocking numbers and the max rated clock speed for the CPU

We’ve talked in the past, extensively, why an extra powerful home only unit is probably a terrible business idea but I think having them both at launch would be extra disastrous.

The perf wouldn’t be great either. 7.5 TFLOPs is the absolute limit, and the CPU perf of an entry level Mac laptop. Everything that makes T239 a remarkable little device makes it a shitty big one

 

[Adulfzen]

So the Switch Lite was a bit of a letdown for Nintendo especially in Japan right ? which would have been it's main target market I imagine. I know this 99% won't happen but imagine if instead of offering a mobile only option for Switch 2 they offered a home console only version and here's the twist, both would be available at launch much like the Series S and Series X.

Just for fun could anyone do the math and see what the maximum clocks, performance and tflops Drake could output with say a 200w power budget instead of sub 20w?

I guess it could go -

Switch 2 Hybrid - 720p mobile / 900p docked then DLSS up to 1440p.
Switch 2 Home - 1080p then DLSS up to 4k. Much more performant Ray Tracing.

$399 for the Hybrid / $499 for the Home (Joy-Cons replaced with Pro Controller 2).

home console version would sell far worse than the Switch Lite ever did and I doubt Nintendo would go for a more powerful sku anyway as it just complicates development rather than just having undocked and docked mode.

 

[Shoulder]

It's not a gimmick, though.

One of the definitions of gimmick:

1. An innovative or unusual mechanical contrivance; a gadget.

Scrolling shoulder buttons sure sounds like a gimmick to me.

 

[RedSpring]

Whatever Nintendo does with the carts, I hope they figure out a shape that is future proof for multiple generations. Adding pins so that it won't fit old devices is easy to understand but it doesn't bode well for multiple generations BC. If I put a PS4 disc in a PS2 it would just give an error and that's how it should be.

 

[Alovon11]

That's why I bring up the ROPs problem before. I feel 16 ROPs isn't gonna cut it.

Actually looked into this with ReddDreadtheLead, and actually 16ROPs may actually be fine. At least for the power-level of the system itself.

DLSS/XeSS/FSR2/Any Gen2 Upscaler like it are Compute-Kernel based, so all they care about is the Internal Resolution of the image being shaded with regards to Pixel Fill Rate (The main thing ROPs influence).


And even then, looking more into how ROPs are defined and the actual pixel-fill rates of TVs...ROPS may have not really been an issue

So, let's use Tegra X1 for example as it has 16ROPs (But is far smaller than T239), with a Pixel Fill Rate of roughly 12GP/s

So Tegra X1 in Switch at its clocks would be able to fill 12000 Megapixels per second. For Reference 4K at 60 is only around 500 Megapixels per second. Heck, you could theoretically pump 4K 120fps content through Tegra X1 and it wouldn't stutter via a Pixel-fill bottleneck.

Now, this is per-RenderTarget...Although doing more research on that

EDIT: and just did some napkin-math on a frame of multiple buffers and 1440p 60 with 22 frametargets/segments, what sounds like a worst case would only fill up 4 Gigapixels per second. And not all of those even engage with the ROPs so that is a vast overestimate

EDIT 2: Okay so Legacy Pipelines are the main thing that depend on ROPs. AKA UE4/Games that prioritize Pixel Shaders rather than compute shaders. So 16ROPs may limit you in games that rely on those pipelines. But games on Drake likely will moreso focus on Compute Shaders (at NVIDIA/Nintendo's recommendation likely) so that sidesteps that even more than DLSS/FSR2/XeSS would.

 

[Hosermess]

So the Switch Lite was a bit of a letdown for Nintendo especially in Japan right ? which would have been it's main target market I imagine. I know this 99% won't happen but imagine if instead of offering a mobile only option for Switch 2 they offered a home console only version and here's the twist, both would be available at launch much like the Series S and Series X.

Just for fun could anyone do the math and see what the maximum clocks, performance and tflops Drake could output with say a 200w power budget instead of sub 20w?

I guess it could go -

Switch 2 Hybrid - 720p mobile / 900p docked then DLSS up to 1440p.
Switch 2 Home - 1080p then DLSS up to 4k. Much more performant Ray Tracing.

$399 for the Hybrid / $499 for the Home (Joy-Cons replaced with Pro Controller 2).

Don't think Switch Lite is much of a letdown in Japan anyway. Think they're ok with people buying the hybrid Switches instead.

 

[RedSpring]

So the Switch Lite was a bit of a letdown for Nintendo especially in Japan right ? which would have been it's main target market I imagine. I know this 99% won't happen but imagine if instead of offering a mobile only option for Switch 2 they offered a home console only version and here's the twist, both would be available at launch much like the Series S and Series X.

Just for fun could anyone do the math and see what the maximum clocks, performance and tflops Drake could output with say a 200w power budget instead of sub 20w?

I guess it could go -

Switch 2 Hybrid - 720p mobile / 900p docked then DLSS up to 1440p.
Switch 2 Home - 1080p then DLSS up to 4k. Much more performant Ray Tracing.

$399 for the Hybrid / $499 for the Home (Joy-Cons replaced with Pro Controller 2).

A stationary console only I can see but a more powerful console would be a return to handheld and console split. Except this time, they have a handheld that cost more due to extra parts and cooling so it can functions as a less appealing hybrid product and a console that is forced to be much weaker instead of by choice. Mobile chips are efficient sure but they also can't clock as high either. This will be more expensive and still be weaker than a series S.

 

[Giancarlo]

A stationary console only I can see but a more powerful console would be a return to handheld and console split. Except this time, they have a handheld that cost more due to extra parts and cooling so it can functions as a less appealing hybrid product and a console that is forced to be much weaker instead of by choice. Mobile chips are efficient sure but they also can't clock as high either. This will be more expensive and still be weaker than a series S.

i dont see why Nintendo cant release two console at ounce, this patern of a home/handheld console always worked for Nintendo, in what you porpuse, developers can easily reduce assest, resolution for the Switch sucessor hybrid model and increase it assest/resolution for the Switch sucessor home console edtion, today games engines are quite scallable, just do some tweaking in your engine, and you can adapt your game for this model/consoles

 

[Dekuman]

i dont see why Nintendo cant release two console at ounce, this patern of a home/handheld console always worked for Nintendo

I can see different form factors under the same platform but hybrid is doing well for them and they are likely to stick with that.

I doubt they will do a scree less home only version. They will do a Lite model instead

 

[Giancarlo]

I can see different form factors under the same platform but hybrid is doing well for them and they are likely to stick with that.

I doubt they will do a scree less home only version. They will do a Lite model instead

Switch successor hybrid model, Switch sucessor OLED edition 2/3 years later

 

[Ojoloco]

Looks like nintendo go to a minorist distributor event

 

[Samuspet]

I know it doesn't make business sense after the runaway Switch success y'all I just thought it would be interesting to know how powerful it could theoretically be. Sorry if it's already been discussed to death it's just I've never seen it be talked about my bad.

@oldpuck

I guess it would be majorly CPU bound then and all for some extra crispiness most people wouldn't notice (1440p vs 2160p DLSS) or higher resolution RT reflections. Not worth the confusing the market! Thanks.

 

[KeijiKG]

Looks like nintendo go to a minorist distributor event

How peculiar.

Voice actor Ashly Burch?
And it seems there was a display for Super Mario Bros. Wonder.

 

[LiC]

Don't think Switch Lite is much of a letdown in Japan anyway. Think they're ok with people buying the hybrid Switches instead.

There's also ground in between a runaway success and a letdown. The Switch Lite was obviously created with the idea that some people who wouldn't buy a hybrid Switch either due to cost, or the complexity of the hardware proposition, would buy a cheaper and simpler handheld (or that some households that already had a Switch would be more likely to buy a cheaper option for their second unit). None of us can say, and even Nintendo with their market data can only estimate, but it's reasonable to conclude that this has been borne out and Nintendo did in fact reach additional customers with the Lite.

 

[ezidro3]

How peculiar.

Voice actor Ashly Burch?
And it seems there was a display for Super Mario Bros. Wonder.

Yeah weird combo. Her with the SMBW shirt makes me think she might have a small voice role in it? Since the trailer had more voiced elements than previous games

 

[Shoulder]

i dont see why Nintendo cant release two console at ounce, this patern of a home/handheld console always worked for Nintendo, in what you porpuse, developers can easily reduce assest, resolution for the Switch sucessor hybrid model and increase it assest/resolution for the Switch sucessor home console edtion, today games engines are quite scallable, just do some tweaking in your engine, and you can adapt your game for this model/consoles

“Never half-ass two things. Whole ass one thing.” - Ron Swanson

What ever Nintendo did back then doesn’t imply it’ll still work today. They got something good here, so I would hope they’re going to continue it from then on.

 

[OctoSplattack]

Even if Nintendo decides to make a non-portable version of the Switch... what kind of niche would that fill for them? As it is, they already have the budget model (Lite), standard model (V2), and premium model (OLED). And even if the Lite can't dock to a TV, all three still stick to the basic idea of allowing home console gaming on-the-go. No matter which one you buy, you can "Play anytime, anywhere, with anyone." I'd assume they'd want to repeat that setup with the Switch NG.

A "Switch Home" would break that idea, even if it could feasibly be an extra-budget model if they just pulled a PS Vita TV and made a microconsole that they can sell with a pair of Joy-Con and a pack-in title for $100-150. I wouldn't rule it out necessarily, but it would mess with branding too much. And making it more powerful doesn't feel like a option on any level, since that's not only a performance profile that no other version of the Switch would be able to make use of, but it would be weird brand messaging for the best performing version of the device to be incapable of the entire hardware line's main selling point of portable play.

 

[kvetcha]

Even if Nintendo decides to make a non-portable version of the Switch... what kind of niche would that fill for them? As it is, they already have the budget model (Lite), standard model (V2), and premium model (OLED). And even if the Lite can't dock to a TV, all three still stick to the basic idea of allowing home console gaming on-the-go. No matter which one you buy, you can "Play anytime, anywhere, with anyone." I'd assume they'd want to repeat that setup with the Switch NG.

A "Switch Home" would break that idea, even if it could feasibly be an extra-budget model if they just pulled a PS Vita TV and made a microconsole that they can sell with a pair of Joy-Con and a pack-in title for $100-150. I wouldn't rule it out necessarily, but it would mess with branding too much. And making it more powerful doesn't feel like a option on any level, since that's not only a performance profile that no other version of the Switch would be able to make use of, but it would be weird brand messaging for the best performing version of the device to be incapable of the entire hardware line's main selling point of portable play.

At best, it feels like a late-generation volume play, but even that isn't really convincing.

 

[fwd-bwd]

I suspect it may be possible to cleanly handle a new card format with the addition of one new pair of pins. […] If the Switch NG cards are a bit wider (I’m assuming there will be some physical change in any case, to prevent them being inserted into the original Switch), then an extra pair of pins could be added next to the existing pins (let’s say pins -1/0 or pins 18/19) to act as new card detect pins.

Whatever Nintendo does with the carts, I hope they figure out a shape that is future proof for multiple generations. Adding pins so that it won’t fit old devices is easy to understand but it doesn’t bode well for multiple generations BC.

If adding two pins is all it needs to a) enable higher speed, and b) differentiate the Game Card generations, I think that it can be achieved without changing the shape of card or even the pin layout.

As the diagram shown, the current pins 5, 8, 11, 16, and 17 each can be bisected into two contacts, potential increasing the total up to 22 pins without changing the physical design of the Game Card, while ensuring backward compatibility. I’m of the opinion that it’d fit Nintendo’s product strategy better to have one physical SKU per cross-gen title, instead of selling two separate cards (Switch 1 and NG) of the same game.

 

[RedSpring]

i dont see why Nintendo cant release two console at ounce, this patern of a home/handheld console always worked for Nintendo, in what you porpuse, developers can easily reduce assest, resolution for the Switch sucessor hybrid model and increase it assest/resolution for the Switch sucessor home console edtion, today games engines are quite scallable, just do some tweaking in your engine, and you can adapt your game for this model/consoles

I would love two consoles from them but I can only see two devices hampering each other going with the series route. You’re still contrained by the hybrid model so if power is keeping games away then having a stronger model won’t fix it. If the games are coming then what audience is the stronger model catering to? Will it makes up for making the hybrid model looks compromised? It will still be noticably weaker without any cost saving. The selling point of the series is you can get a cheap next gen console or pay extra for a top of the line hardware. This switch 2 will be that you can pay a premium for a handheld that dock but the power user can pay more for a weak console. It might be top of the line mobile hardware sure but it’s in a stationary console. Not the same marketing appeal.

 

[ILikeFeet]

If adding two pins is all it needs to a) enable higher speed, and b) differentiate the Game Card generations, I think that it can be achieved without changing the shape of card or even the pin layout.

As the diagram shown, the current pins 5, 8, 11, 16, and 17 each can be bisected into two contacts, potential increasing the total up to 22 pins without changing the physical design of the Game Card, while ensuring backward compatibility. I’m of the opinion that it’d fit Nintendo’s product strategy better to have one physical SKU per cross-gen title, instead of selling two separate cards (Switch 1 and NG) of the same game.

does the size of the pins affect anything or is it somewhat arbitrary

 

[Shoulder]

Even if Nintendo decides to make a non-portable version of the Switch... what kind of niche would that fill for them? As it is, they already have the budget model (Lite), standard model (V2), and premium model (OLED). And even if the Lite can't dock to a TV, all three still stick to the basic idea of allowing home console gaming on-the-go. No matter which one you buy, you can "Play anytime, anywhere, with anyone." I'd assume they'd want to repeat that setup with the Switch NG.

A "Switch Home" would break that idea, even if it could feasibly be an extra-budget model if they just pulled a PS Vita TV and made a microconsole that they can sell with a pair of Joy-Con and a pack-in title for $100-150. I wouldn't rule it out necessarily, but it would mess with branding too much. And making it more powerful doesn't feel like a option on any level, since that's not only a performance profile that no other version of the Switch would be able to make use of, but it would be weird brand messaging for the best performing version of the device to be incapable of the entire hardware line's main selling point of portable play.

And let’s face it, their consoles haven’t been as successful overall compared to their handhelds, and especially these days, what purpose does it serve to continue down that road again?

For the sake of argument, let’s do some averages between their consoles vs. handhelds.

Consoles:
NES - 61.91 million
SNES - 49.10 million
N64 - 32.93 million
GCN - 21.74 million
Wii - 101.63 million
Wii U - 13.56 million

Handhelds
Game & Watch - 43.4 million
GB/GBC - 118.69 million
GBA - 81.51 million
DS - 154.02 million
3DS - 75.94 million

Now let’s take their combined totals.

All Consoles: 280.87 million
All Handhelds: 473.56 million

Average for Consoles: 46.81 million

Average for Handhelds: 94.71 million

Let’s take it a step further though, and remove the two lowest selling systems for consoles, and handheld, and recalculate.

Total Consoles: 267.31 million
Total Handhelds: 430.16 million

Take those, and our new averages are:

Avg Consoles: 53.46 million
Avg Handheld: 107.54 million

Now how about this then. They’ve made six consoles prior to Switch, but only five handhelds. So let’s assume they only made 5 consoles, same as Handhelds, but made six handhelds, and calculate the original figures again

Avg Console (assuming five systems): 56.17 million

Avg Handhelds (assuming six handhelds): 78.92 million

No matter how you slice it (and there are other ways to calculate some median numbers here), their handhelds have vastly overall been more successful than their consoles. We’re talking a 2:1 difference in lifetime sales. You can very much argue some aspects of the data I’m sure, plus the fact that Nintendo has practically a monopoly when it comes to gaming-based handhelds, compared to stiff competition with their consoles.

At the same time though, it’s been proven over time their handhelds have stood the test of time even in the face of competition from Sega, and Sony, the latter being the most successful of the two competitors, yet still could not topple Nintendo.

The Switch is really a best of both worlds system, and even for me is difficult to admit sometimes considering I was totally against the idea of a single hybrid console. @Goodtwin for example knows me long enough that I had doubts this hybrid system would end up successful.

Ultimately, I feel Nintendo would be doing themselves a disservice going back to one handheld, and one console, plus the fanbase for that matter. Just my take on it.

EDIT:

Just to add to it, let’s remove the highest selling platforms of each, and recalculate again:

Consoles: 179.24 million
Handhelds: 319.54 million

Avg Console: 35.48 million
Avg Handheld: 79.88 million.

Still 2:1, actually 2.2:1, so it fairs worse in this scenario.

Yeah. Handhelds for Nintendo just sell better. Simple as that.

 

[fwd-bwd]

does the size of the pins affect anything or is it somewhat arbitrary

Not mechanically, as you see from the photo of card slot:

It seems to me that the length of each pin on the Game Card is used to indicate its function for human readability. For examples, pins 8 and 11 are longer because they are VCC inputs, and 16 the longest to denote it being the GND.

 

[ILikeFeet]

Not mechanically, as you see from the photo of card slot:

It seems to me that the length of each pin on the Game Card is used to indicate its function for human readability. For examples, pins 8 and 11 are longer because they are VCC inputs, and 16 the longest to denote it being the GND.

then it looks like the solution was right there the whole time

 

[RedSpring]

If adding two pins is all it needs to a) enable higher speed, and b) differentiate the Game Card generations, I think that it can be achieved without changing the shape of card or even the pin layout.

That's good to know. Hopefully Nintendo don't keep on changing the shape of the card every generations like they have been doing because they think it would confuse people that they can insert the card but not play it. If people can figure out disc, they can figure out card.

 

[Deleted member 887]

I know it doesn't make business sense after the runaway Switch success y'all I just thought it would be interesting to know how powerful it could theoretically be. Sorry if it's already been discussed to death it's just I've never seen it be talked about my bad.

Nah, no apologies! It's an interesting question. Two power profiles obviously works - with the PS5 Pro coming, it seems like it'll happen two gens in a row for both MS and Sony. But because the Switch already has two profiles, a more powerful TV only unit opens it up to three. I think it actually is viable, just probably not at launch.

@oldpuck

I guess it would be majorly CPU bound then and all for some extra crispiness most people wouldn't notice (1440p vs 2160p DLSS) or higher resolution RT reflections. Not worth the confusing the market! Thanks.

I mean, I'm willing to bet that overclockers will go there as soon as they can .

In the case of this particular chip, it's very big... for a tablet. There is an area of very low clocks, between 300 and 500 MHz, way below where Nvidia would usually clock a GPU, that is super power efficient. By having way more cores than even the Steam Deck has, Drake can get a nice amount of performance out of minimum electricity - low clocks, big GPU.

But it's very small for a desktop GPU. So even if you do throw mondo electricity at it and put it to absolute max clocks, it's basically an overclocked laptop.

 

[TLZ]

“Never half-ass two things. Whole ass one thing.” - Ron Swanson

What ever Nintendo did back then doesn’t imply it’ll still work today. They got something good here, so I would hope they’re going to continue it from then on.

The Switch is the definition of "half-ass two things". It's neither a full on home console that isn't limited by low wattage and small space, nor a full on portable that's pocketable like they used to be.

It's a jack of all trades. And it's working for them.

 

[7hatDeadCat]

Nah, no apologies! It's an interesting question. Two power profiles obviously works - with the PS5 Pro coming, it seems like it'll happen two gens in a row for both MS and Sony. But because the Switch already has two profiles, a more powerful TV only unit opens it up to three. I think it actually is viable, just probably not at launch.


I mean, I'm willing to bet that overclockers will go there as soon as they can .

In the case of this particular chip, it's very big... for a tablet. There is an area of very low clocks, between 300 and 500 MHz, way below where Nvidia would usually clock a GPU, that is super power efficient. By having way more cores than even the Steam Deck has, Drake can get a nice amount of performance out of minimum electricity - low clocks, big GPU.

But it's very small for a desktop GPU. So even if you do throw mondo electricity at it and put it to absolute max clocks, it's basically an overclocked laptop.

Why would it open it up to 3 profiles? Do you mean more like 3 SKUs? The TV only, which would run at the docked performance mode. the Hybrid, which can do docked performance mode and portable performance mode. And the Lite, which is limited to portable mode.

 

[OctoSplattack]

Why would it open it up to 3 profiles? Do you mean more like 3 SKUs? The TV only, which would run at the docked performance mode. the Hybrid, which can do docked performance mode and portable performance mode. And the Lite, which is limited to portable mode.

Three profiles because the person they were replying to was proposing that the TV-only be more powerful than the hybrid.

 

[7hatDeadCat]

Three profiles because the person they were replying to was proposing that the TV-only be more powerful than the hybrid.

That makes sense, dang my skimming brain i passed right over that.

 

[Deleted member 887]

Why would it open it up to 3 profiles? Do you mean more like 3 SKUs? The TV only, which would run at the docked performance mode. the Hybrid, which can do docked performance mode and portable performance mode. And the Lite, which is limited to portable mode.

Samuspet was specifically referring to was a TV only device that was more expensive and more powerful than just docked mode.