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

[Brofield]

Fortunately, marketing decisions are much more flexible and fluid than hardware decisions.

~



Jon from Spawn Wave was doing a quick summary of what NateDrake and MVG said on the latest Nate the Hate episode released yesterday for the most part. However, Jon brought up a question near the end of the video, asking would there be compatibility on the Nintendo Switch via the cloud, where a game runs natively on the DLSS model*, but consumers can still play that game on the Nintendo Switch by buying the cloud version of that game? Jon mentions Hitman 3, Control, and Kingdom Hearts as examples.

So the question I have is will third party developers bother releasing cross-gen games by having games run natively on the DLSS model* and having games run via the cloud on the Nintendo Switch?

To be honest I had a similar train of thought a while back, the idea that it would most likely be effective for 3rd parties who already have PC ports of games, and buying the cloud versions would also eventually translate to a digital code of the game in question once the user upgrades.

Really, I think the only thing that prevents this idea from flourishing is people on the internet shouting down those who suggest it because they have poor internet connections to begin with and don't want to see it become the standard.

 

[Skittzo]

It doesnt even cross my mind that forward compatibility wasnt a factor when designing the switch. They're kinda forced with the way they've been pushing digital sales, it would be beyond the pale not to do it. And bc has always been a priority for nintendo whenever possible

Yeah I absolutely agree. I'm sure they've worked out a way to future proof the library, it'll just have to be slightly less straightforward than native binary compatibility.

 

[fwd-bwd]

Jon from Spawn Wave was doing a quick summary of what NateDrake and MVG said on the latest Nate the Hate episode released yesterday for the most part. However, Jon brought up a question near the end of the video, asking would there be compatibility on the Nintendo Switch via the cloud, where a game runs natively on the DLSS model*, but consumers can still play that game on the Nintendo Switch by buying the cloud version of that game? Jon mentions Hitman 3, Control, and Kingdom Hearts as examples.

So the question I have is will third party developers bother releasing cross-gen games by having games run natively on the DLSS model* and having games run via the cloud on the Nintendo Switch?

While technically feasible, from the product management standpoint the business case seems weak.

• Since the cloud version (most likely a Windows game) can work on either Switch 1 or Switch 2, there's little incentive to spend more to develop a Switch 2 version.
• Cloud gaming infrastructure and ongoing operational cost are not cheap. The dev would want more users on the cloud for a better ROI. Why splintering your user base between the cloud and Switch 2?
• As @OtterTrophic pointed out, having two competing versions may be rather difficult to market. User confusion probably would be unavoidable.

For Nintendo, I'd think that they prefer developers releasing native games on Switch instead of streaming PC games to Switch. A potential solution is to host Switch 2 games on ARM servers running Switch 2 virtualization, and stream to any Switch (1 or 2) with an NSO subscription. This model would encourage devs to develop actual games for the Switch hardware, as opposed to putting PC games on the cloud for the Switch market. The dev receives revenues from selling their games to Switch 2 owners and also from streaming to NSO subscribers; the cost of cloud operations essentially is subsidized by Nintendo via NSO.

 

[ILikeFeet]

While technically feasible, from the product management standpoint the business case seems weak.

• Since the cloud version (most likely a Windows game) can work on either Switch 1 or Switch 2, there's little incentive to spend more to develop a Switch 2 version.
• Cloud gaming infrastructure and ongoing operational cost are not cheap. The dev would want more users on the cloud for a better ROI. Why splintering your user base between the cloud and Switch 2?
• As @OtterTrophic pointed out, having two competing versions may be rather difficult to market. User confusion probably would be unavoidable.

For Nintendo, I'd think that they prefer developers releasing native games on Switch instead of streaming PC games to Switch. A potential solution is to host Switch 2 games on ARM servers running Switch 2 virtualization, and stream to any Switch (1 or 2) with an NSO subscription. This model would encourage devs to develop actual games for the Switch hardware, as opposed to putting PC games on the cloud for the Switch market. The dev receives revenues from selling their games to Switch 2 owners and also from streaming to NSO subscribers; the cost of cloud operations essentially is subsidized by Nintendo via NSO.

where will Nintendo get that infrastructure from? Nvidia's Geforce Now is running Windows servers with Intel CPUs

 

[fwd-bwd]

where will Nintendo get that infrastructure from? Nvidia's Geforce Now is running Windows servers with Intel CPUs

Yes, GFN is Intel based today. My speculation is about the future. Nvidia has serious ambition for the ARM server business, and their attempt to acquire ARM exemplifies that. A partnership with Nintendo to build an NSO cloud seems within the realms of possibility.

 

[Dakhil]

where will Nintendo get that infrastructure from? Nvidia's Geforce Now is running Windows servers with Intel CPUs

Jensen Huang mentioned at 41:33 in the GTC 2021 (April 2021) keynote that GeForce Now is transitioning to BlueField.



Edit: Apparently, Mediatek has released mobile ray tracing extensions for Vulkan, and has partnered with Arm and Tencent Games to showcase real-time ray tracing technology on mobile devices.

 

[Adahel]

Jensen Huang mentioned at 41:33 in the GTC 2021 (April 2021) keynote that GeForce Now is transitioning to BlueField.



Edit: Apparently, Mediatek has released mobile ray tracing extensions for Vulkan, and has partnered with Arm and Tencent Games to showcase real-time ray tracing technology on mobile devices.

Have we a release date confirmation for the ARM+NV GPU? It would be great to have some performance data (aside for Nvidia provided) and see how it's handled on different workloads

 

[Helix]

give the very nature of the speculated hardware we are seeing here, it just seems off for Nintendo to just call the Dane a mid-cycle refresh.

I know we are more fluid when it comes to generational leaps these days but naming this device anything that isn't indicative of a significant jump from the Switch kinda feels wrong now.

 

[Intoxicate]

With the OLED update two years after V2 (which again came 2 years after the V1) I would assume Nintendo is closely monitoring the sales and if it’s still on the same level I would expect an OLED Switch lite in 2022 first prior to another dockable Switch update.

Regarding BC; isn’t Nintendo’s strategy still to increase the installed base, and thriving for an iOS like environment?

 

[YolkFolk]

give the very nature of the speculated hardware we are seeing here, it just seems off for Nintendo to just call it the Dane a mid-cycle refresh.

I know we are more fluid when it comes to generational leaps these days but naming this device anything that isn't indicative of a significant jump from the Switch kinda feels wrong now.

It’s not a mid cycle refresh. It’s the Switch successor.

 

[Sam Albro]

If BC is being called into question because the chip doesn't natively run code written for the Switch chip, then this isn't a mid-gen refresh. It would have to be a new generation.

 

[My Tulpa]

Based on reports the Switch 4K seems to be a full successor.

Tech wise? Sure. But by Nintendo? Nah.

Nintendo could have released Game Boy Color as a true successor to their original handheld, but they didn’t.

Nintendo will treat this next model as a revision/iteration but still a part of the Switch family.

The OLED Switch was produced and released to extend the life of the OG/base Switch. They didn’t release the $350 OLED to essentially be a dead in product ~1 year later.

The OLED Switch is meant to replace the 2017/2019 Switch models and last another 4-5 years.

 

[My Tulpa]

It’s not a mid cycle refresh. It’s the Switch successor.

It’s a mid cycle refresh in that Nintendo still expects to make games that run on the 2017 Switch for another 4-5 years.

I guarantee you that most Nintendo games released in 2025 will be designed to run on the OLED Switch people bought in 2021

 

[ITSMILNER]

Another thing with BC, if they do have issues in how it would work does this also shoot down any potential for a system level 'Boost mode' on the new Switch for old software?

 

[ILikeFeet]

With the OLED update two years after V2 (which again came 2 years after the V1) I would assume Nintendo is closely monitoring the sales and if it’s still on the same level I would expect an OLED Switch lite in 2022 first prior to another dockable Switch update.

You need to sign contracts to make parts for these devices far in advance. Nintendo can't pivot like this. They're locked in, sales be damned

 

[Onilink]

While technically feasible, from the product management standpoint the business case seems weak.

• Since the cloud version (most likely a Windows game) can work on either Switch 1 or Switch 2, there's little incentive to spend more to develop a Switch 2 version.
• Cloud gaming infrastructure and ongoing operational cost are not cheap. The dev would want more users on the cloud for a better ROI. Why splintering your user base between the cloud and Switch 2?
• As @OtterTrophic pointed out, having two competing versions may be rather difficult to market. User confusion probably would be unavoidable.

For Nintendo, I'd think that they prefer developers releasing native games on Switch instead of streaming PC games to Switch. A potential solution is to host Switch 2 games on ARM servers running Switch 2 virtualization, and stream to any Switch (1 or 2) with an NSO subscription. This model would encourage devs to develop actual games for the Switch hardware, as opposed to putting PC games on the cloud for the Switch market. The dev receives revenues from selling their games to Switch 2 owners and also from streaming to NSO subscribers; the cost of cloud operations essentially is subsidized by Nintendo via NSO.

There is a giant flaw in this theory:
If you (Nintendo) want to suicide yourself and cut out the previous install base, by selling again your games, at this point go to the remaster way, there is no point to made a cloud game version.

 

[YolkFolk]

It’s a mid cycle refresh in that Nintendo still expects to make games that run on the 2017 Switch for another 4-5 years.

I guarantee you that most Nintendo games released in 2025 will be designed to run on the OLED Switch people bought in 2021

Its not a mid cycle refresh. Clearly it isn’t.

Neither is the PS5 for that matter and they have the biggest Sony games coming to PS5 and PS4 still.

I’m sure there will be a few games on the first Switch for a while but that’s normal when moving across into new generations these days.

 

[RainTech]

Yay first post!

I'd say Nintendo is very likely doing BC with the code running natively on the CPU side, while the GPU gets emulated. Nintendo already experimented with that "Half emulation" approach on Mario Galaxy from the 3D Collection. I'm no engineer or any kind of expert, but i hope Nvidia and Nintendo find a way to implement DLSS into BC. I wouldn't be mad if they didn't but it would be a nice bonus.

 

[Hermii]

Yay first post!

I'd say Nintendo is very likely doing BC with the code running natively on the CPU side, while the GPU gets emulated. Nintendo already experimented with that "Half emulation" approach on Mario Galaxy from the 3D Collection. I'm no engineer or any kind of expert, but i hope Nvidia and Nintendo find a way to implement DLSS into BC. I wouldn't be mad if they didn't but it would be a nice bonus.

Dont get your hopes up for that, I think its pretty much impossible. Not an engineer either.

I think we should all be very happy if unpatched software runs at maximum dynamic res/ fps at most times and has better loading times. That should be the best case scenario.

 

[Ryosuke]

Not gonna lie, just the speculation that the next console might not be BC already changed some of my purchase habits / made me regret some purchases that I made.

I'll probably start skipping sales of games that don't run so well, because before I had the mindset "At least the next one might run this fine", which is dangerous I know but considering how sporadic some games go on sale, I didn't mind. Now that's over, for sure.

Giving users a cloud version of past games would be just a slap on the face, I'm not at all a fan of game streaming and don't plan on using that anytime soon.

--

Edited for clarity.

 

[Skittzo]

With the OLED update two years after V2 (which again came 2 years after the V1) I would assume Nintendo is closely monitoring the sales and if it’s still on the same level I would expect an OLED Switch lite in 2022 first prior to another dockable Switch update.

Regarding BC; isn’t Nintendo’s strategy still to increase the installed base, and thriving for an iOS like environment?

I'm pretty confident in thinking an OLED Lite will not be happening, period. The purpose of the OLED model is to get Lite owners and primarily handheld users to upgrade to a more premium product. Introducing an OLED Lite version completely destroys that purpose.

Not gonna lie, just the rumor that the next console might not be BC already changed some of my purchase habits / made me regret some purchases that I made.

I'll probably start skipping sales of games that don't run so well, because before I had the mindset "At least the next one might run this fine", which is dangerous I know but considering how sporadic some games go on sale, I didn't mind. Now that's over, for sure.

Giving users a cloud version of past games would be just a slap on the face, I'm not at all a fan of game streaming and don't plan on using that anytime soon.

Again, there is no rumor to that effect. I don't get how this has been so misunderstood.

Nobody has said the next console won't have BC, there have been literally zero credible rumors from anybody reasonable to that effect. What has been said is that BC will be less straightforward than we think from Nintendo's and Nvidia's perspective, but again this is not a rumor and is simply based on the known technology.

 

[Ryosuke]

Again, there is no rumor to that effect. I don't get how this has been so misunderstood.

Nobody has said the next console won't have BC, there have been literally zero credible rumors from anybody reasonable to that effect. What has been said is that BC will be less straightforward than we think from Nintendo's and Nvidia's perspective, but again this is not a rumor and is simply based on the known technology.

Sorry but that to me is just splitting hairs.
Do we know what the next Switch will be, period? Is the actual CPU/GPU/APU out? Until the actual thing leaks that's all speculation, thus a rumor to me.

I can change what I wrote to "just the speculation that it might not have because of what people are talking", the end result is still the same to me, I won't take risks.

 

[Skittzo]

Sorry but that to me is just splitting hairs.
Do we know what the next Switch will be, period? Is the actual CPU/GPU/APU out? Until the actual thing leaks that's all speculation, thus a rumor to me.

Huh?

What exactly are you thinking is a rumor here, the chip being used in the next hardware or the discussion about BC?


The BC discussion has been solely based on current Switch developers and dataminers discussing how Switch games are compiled and run on the current hardware, and that any new Switch or successor that uses a newer architecture cannot be natively compatible with existing Switch games due to the way drivers are compiled in each Switch game file. Thus Nintendo and Nvidia need to work out some other solution to ensure they are backwards compatible on whatever comes next.

None of that is based on any rumor whatsoever, it's all based on officially known fact.

 

[lemonfresh]

Not gonna lie, just the rumor that the next console might not be BC already changed some of my purchase habits / made me regret some purchases that I made.

I'll probably start skipping sales of games that don't run so well, because before I had the mindset "At least the next one might run this fine", which is dangerous I know but considering how sporadic some games go on sale, I didn't mind. Now that's over, for sure.

Giving users a cloud version of past games would be just a slap on the face, I'm not at all a fan of game streaming and don't plan on using that anytime soon.

Honestly if it doesn't you can just keep your old switch, you don't have to sell it after the new one comes out. I almost never sell my old consoles.

 

[Ryosuke]

Huh?

What exactly are you thinking is a rumor here, the chip being used in the next hardware or the discussion about BC?


The BC discussion has been solely based on current Switch developers and dataminers discussing how Switch games are compiled and run on the current hardware, and that any new Switch or successor that uses a newer architecture cannot be natively compatible with existing Switch games due to the way drivers are compiled in each Switch game file. Thus Nintendo and Nvidia need to work out some other solution to ensure they are backwards compatible on whatever comes next.

None of that is based on any rumor whatsoever, it's all based on officially known fact.

I'm talking about the end. To have BC, Nvidia and Nintendo will have to work out some solution, so the base status is that it won't have anything natively.

What I'm saying is that I am not expecting them to add anything until someone shows me that the thing has BC.

 

[Skittzo]

I'm talking about the end. To have BC, Nvidia and Nintendo will have to work out some solution, so the base status is that it won't have anything natively.

What I'm saying is that I am not expecting them to add anything until someone shows me that the thing has BC.

Okay all I was doing then was clarifying that this is not a rumor. There is nobody suggesting the next thing won't have BC. Just a discussion that it will not be a simple matter.

If you don't expect BC then yeah, act accordingly. Personally I have zero doubt in my mind that it will have BC.

 

[ILikeFeet]

Not gonna lie, just the rumor that the next console might not be BC already changed some of my purchase habits / made me regret some purchases that I made.

I'll probably start skipping sales of games that don't run so well, because before I had the mindset "At least the next one might run this fine", which is dangerous I know but considering how sporadic some games go on sale, I didn't mind. Now that's over, for sure.

Giving users a cloud version of past games would be just a slap on the face, I'm not at all a fan of game streaming and don't plan on using that anytime soon.

there is no rumor. just certain people blowing something out of proportion

 

[RennanNT]

To have BC, Nvidia and Nintendo will have to work out some solution, so the base status is that it won't have anything natively.

Just wanted to clarify the situation as well.

Every time Nintendo did BC, they paid a price for it. Either by using outdated technology, or adding components from the previous console or making a Frankenstein chip.

The latest ARM CPU cores are still compatible with the CPU instructions used by Switch games, the games aren't tied to a specific hardware and ARM is still the best architecture for a mobile chip. So people were expecting a cheaper than ever BC.

The driver being within the game gave them a little more performance at the cost of BC becoming harder, but they still have their old ways to achieve BC and better new ways as well to chose from, they just don't have the cheapest one that people were expecting for.

 

[ReddDreadtheLead]

Just to note, the issue that the Dane unit faced is similar to that the One X and PS4 Pro faced, and more so that PS5 and the Series faced too.

Sorry but that to me is just splitting hairs.
Do we know what the next Switch will be, period? Is the actual CPU/GPU/APU out? Until the actual thing leaks that's all speculation, thus a rumor to me.

I can change what I wrote to "just the speculation that it might not have because of what people are talking", the end result is still the same to me, I won't take risks.

What type of spin is this

 

[Ryosuke]

Just to note, the issue that the Dane unit faced is similar to that the One X and PS4 Pro faced, and more so that PS5 and the Series faced too.

What type of spin is this

What spin? What are you even talking about.

 

[fwd-bwd]

There is a giant flaw in this theory:
If you (Nintendo) want to suicide yourself and cut out the previous install base, by selling again your games, at this point go to the remaster way, there is no point to made a cloud game version.

That's good argument. The business model of cloud gaming is still evolving. The current thinking is that a subscription model, such as the NSO, can better support the operational costs of the cloud. It provides a recurring revenue for existing titles, but does not preclude the developer from releasing new games/remasters/collections. Take the Super Mario 64 for example, Nintendo is going to start collecting a monthly fee for the original SM64 via the NSO "expansion pack", and they also sold the remastered version as part of the 3D All-Starts. So, if done right, cloud gaming actually may increase choices for the gamers, give old titles an extended life, and generate additional revenues for the developers.

 

[Deleted member 887]

Another thing with BC, if they do have issues in how it would work does this also shoot down any potential for a system level 'Boost mode' on the new Switch for old software?

Improved performance for Classic Switch Games is still on the table.

 

[Deleted member 887]

So the question I have is will third party developers bother releasing cross-gen games by having games run natively on the DLSS model* and having games run via the cloud on the Nintendo Switch?

Been pondering this question. I would put money to say this will happen for at least a couple titles, not sure if that'll be the broad move.

One of the interesting things about cloud gaming is not how it allows "ports" to underpowered hardware, but how it allows last gen hardware to play current gen games. If cloud gaming had existed in the Wii era, with its HUGE install base and users reluctant to move to the successor, you can selling "cloud versions" of games for the Wii forever. If you're using a single unified cloud client to amortize the costs, and a shared cloud backend for all cloud customers it becomes very cheap to add systems supported by your cloud. If you're going to do a cloud release anywhere, do it everywhere.

You could imagine classic Switch getting cloud games for a decade, well beyond what we think of as the "cross-gen" period.

 

[Onilink]

That's good argument. The business model of cloud gaming is still evolving. The current thinking is that a subscription model, such as the NSO, can better support the operational costs of the cloud. It provides a recurring revenue for existing titles, but does not preclude the developer from releasing new games/remasters/collections. Take the Super Mario 64 for example, Nintendo is going to start collecting a monthly fee for the original SM64 via the NSO "expansion pack", and they also sold the remastered version as part of the 3D All-Starts. So, if done right, cloud gaming actually may increase choices for the gamers, give old titles an extended life, and generate additional revenues for the developers.

For SM64, i think that the NSO rom is the same sold in the Mario All Star Collections with NERDS emulator. Of course developers will sell 4k remastered version of games (think only the cloud version of games OR the Doom porting). But it's more a problem about throw away the current userbase and starting from begin, it's insane.

 

[fwd-bwd]

The Bluetooth registration for the N64 Controller HAC-043 is now public (link below). Nothing interesting here, aside from revealing the Bluetooth module being Hosiden HRM1085.

Launch Studio - Listing Details

 

[My Tulpa]

Its not a mid cycle refresh. Clearly it isn’t.

Neither is the PS5 for that matter and they have the biggest Sony games coming to PS5 and PS4 still.

I’m sure there will be a few games on the first Switch for a while but that’s normal when moving across into new generations these days.

I meant, according to Nintendo it will be a revision/iteration.

It will have exclusive games like the Gameboy Color and n3ds did, sure. But most games Nintendo releases in 2025/2026 will still be playable on the OLED Switch people bought in 2021/2022. Of that I’m sure of.

I also don’t see the rtx 3080 series as a “successor” to anything either, tbf.

 

[YolkFolk]

I meant, according to Nintendo it will be a revision/iteration.

It will have exclusive games like the Gameboy Color and n3ds did, sure. But most games Nintendo releases in 2025/2026 will still be playable on the OLED Switch people bought in 2021/2022. Of that I’m sure of.

I also don’t see the rtx 3080 series as a “successor” to anything either, tbf.

I don’t agree. I believe it will be Switch 2.

The hardware is too big of an upgrade. It has brand new custom chip. They’ll want mega sales from this investment and not just from those wanting to upgrade for higher res and better frame rates. When iterations come out, the majority of people stick with the original hardware. Not so with next generation devices. Nintendo will want it known that this hardware is the future of Switch

 

[Pokemaniac]

Tech wise? Sure. But by Nintendo? Nah.

Nintendo could have released Game Boy Color as a true successor to their original handheld, but they didn’t.

Nintendo will treat this next model as a revision/iteration but still a part of the Switch family.

The OLED Switch was produced and released to extend the life of the OG/base Switch. They didn’t release the $350 OLED to essentially be a dead in product ~1 year later.

The OLED Switch is meant to replace the 2017/2019 Switch models and last another 4-5 years.

The Game Boy Color is a completely different class of upgrade than what's being expected with Switch 4k. The thing was basically just an overclocked Game Boy with some graphics upgrades to support color, and was almost certainly only ever made because the GBA ran into some development issues. Switch 4k is looking to move the CPU and GPU to designs multiple generations newer.

Consoles don't just drop dead when their successor releases, especially not recently.

 

[ReddDreadtheLead]

What spin? What are you even talking about.

You’re spinning a misinterpretation of a comment as a fact and when told that it’s not the end of the world it can be implemented, and considering that the company history offers BC for their devices, you call that BC feature a “rumor” and go as far as to imply that it’s speculation that it will have BC.


You’re basically inducing FUD for the sake of it and closing off any other comments on it.

Dane wouldn’t even be the only system to experience this BC issue, it’s happened before and will happen for consoles until they figure out how to actually have a software feature that just works with updates, or they become PCs where it isn’t specific, it’s generalized in coding for, but you won’t get the best out of the hardware.

 

[bmfrosty]

I don’t agree. I believe it will be Switch 2.

The hardware is too big of an upgrade. It has brand new custom chip. They’ll want mega sales from this investment and not just from those wanting to upgrade for higher res and better frame rates. When iterations come out, the majority of people stick with the original hardware. Not so with next generation devices. Nintendo will want it known that this hardware is the future of Switch

It doesn't matter if it's a called a revision or a new console. It's going to be the same thing either way, unless they scrap Dane and start over.

 

[Skittzo]

Consoles don't just drop dead when their successor releases, especially not recently.

Tell that to the PS4 and XB1 sales.

 

[lemonfresh]

Tell that to the PS4 and XB1 sales.

Better yet the wii and wii u

 

[Skittzo]

Better yet the wii and wii u

Eh, the Wii dropped dead even before the successor released. Wii U dropped dead as soon as itself was released.

 

[Pokemaniac]

Tell that to the PS4 and XB1 sales.

I was referring more to the games than the hardware itself, but that's fair. Still, it's probably the Switch Lite that's really going to stick around long term. There's been a fair amount of churn with the hybrid models.

Better yet the wii and wii u

That's more the 3DS's fault than anything else.

 

[YolkFolk]

It doesn't matter if it's a called a revision or a new console. It's going to be the same thing either way, unless they scrap Dane and start over.

It matters a lot.

You think the PS5 would be getting mega sales if it was called the PS4 Pro Plus?

 

[bmfrosty]

It matters a lot.

You think the PS5 would be getting mega sales if it was called the PS4 Pro Plus?

That's not what I'm talking about. It's technologically the same beast either way it's marketed. You might want to start a "Future Nintendo Hardware Marketing Discussion" thread if you want to kibitz over how it's marketed as how it's marketed has very little to do with what technology is in it.

This isn't meant to be dismissive, but this is the third (fourth?) one of these threads and the argument has cropped up continuously in the threads and unlike the rumors on technology only ends up in argument about what marketing approach would be better, which isn't fun for those just wanting to discuss the possible technologies used and trying to sleuth out what technologies will be in future Nintendo products.

 

[ShadowFox08]

I meant, according to Nintendo it will be a revision/iteration.

It will have exclusive games like the Gameboy Color and n3ds did, sure. But most games Nintendo releases in 2025/2026 will still be playable on the OLED Switch people bought in 2021/2022. Of that I’m sure of.

I also don’t see the rtx 3080 series as a “successor” to anything either, tbf.

Huh, Nintendo hasn't confirmed anything.
And Nintendo using Nvidia ampere or better with A78 CPUs would be next generation, not a revision.

But they can't really do anything more with TX1 anyway. They can't over lock GPU by 50% and slap 4 more A57 cores and double or 1.5x CPU speeds and call it a NEW Switch. That's not happening.

 

[BlackTangMaster]

The weakest graphic card with DLSS is still way more powerful than what Dane will be, so we have no real way to say for certain, but the general expectation is that the power saving from lowering native resolution (and thus GPU clock) more than made up for the DLSS cost. But even if that is true, we don't even know if the lower clock would be able to run DLSS targeting 720p.

Also, 540p without DLSS cost less than 540p with DLSS. Devs concerned about battery life may chose to simply not use it to get more savings.

While RTX2060 is obviously more powerful than Orin or Dane could be, the innovation comes from how powerful the tensor cores would be compared to those on TU106. Each Dane tensor cores could be 2* more powerful than TU106 tensor cores and the tensor core count could be doubled compared to Desktop Ampere (and more in line with GA100 which has 2* the amount of Desktop ampere tensor cores at the cost of not having RT cores). Dane could benefit from reusing the GA100 HPC GPU uarch than using the desktop ampere uarch.

I don't know if it is possible to deactivate tensor cores in handheld mode as they make part of the SM instead of being an independent core on the SoC with its own power rail.

These are quite interesting. It's strange that TSMC is ramping up N5 slowly given that it's supposed to be a long-term node and demand will explode with Zen 4, RDNA 3, Lovelace, Apple MX/Axx, and others. N7 will keep being used for the foreseeable future. I also hope N3 isn't a 20nm repeat. They're stretching FinFET to the limits.
I have to say though that these tweets show that Samsung 8N isn't really a bad choice for next Switch SoC. Unless you want the best PPA possible, you don't choose the node based on that alone. Other things need to be assessed carefully. Transistors costs going up instead of down is something the industry has been telling us for a long time.

IMHO, TSMC ramping up is highly tied to Apple investment in TSMC for high end nodes. N5 is undoubtedly the best node available (compared to Intel 7 nm (10SFF) and 5LPP) and Apple don't need to invest more $ as they have most of the best node capacities for 2 to 3 years.

Upcoming talk at Nvidia GTC on Wednesday Nov 10 at 7AM Pacific Time:

NVIDIA AGX Platform: Scalable and Modular Architecture for High Performance AI Compute​

​

This session will provide an in-depth view into the NVIDIA AGX product line up based on the NVIDIA Orin SoC. It will feature the modular and scalable architecture of these products and dive deeper into how these are used across multiple industries such as autonomous vehicles, robotics, and healthcare.​

"Product lineup" and "multiple industries" caught my eyes. "Gaming" is not listed though.

Modular and scalable enough for Switch 2. One day we will probably see Switch OS on a Mercedes car or a Siemens CTscan/Ultrasound machine.

So the question I have is will third party developers bother releasing cross-gen games by having games run natively on the DLSS model* and having games run via the cloud on the Nintendo Switch?

Mon choux tu es mignon (the business case would be worse than before the release of the cloud version).

Tech wise? Sure. But by Nintendo? Nah.

Nintendo could have released Game Boy Color as a true successor to their original handheld, but they didn’t.

Nintendo will treat this next model as a revision/iteration but still a part of the Switch family.

The OLED Switch was produced and released to extend the life of the OG/base Switch. They didn’t release the $350 OLED to essentially be a dead in product ~1 year later.

The OLED Switch is meant to replace the 2017/2019 Switch models and last another 4-5 years.

Actually, I believe that the OLED model was really made with the aim to extend the Switch platform life by 4-5 years but not with the TX1 in it. While PS5 and XSX have probably benefited from contacts made before the covid induced shortages, they will probably be facing a price cost increase that will be higher than originally expected. They may not be able to make these devices profitable before 2 to 3 years. As opposed to Nintendo which are consolidating their margins with OLED while making economy of scale with key components such as the screen and the manufacturing of the new body and dock.

 

[ILikeFeet]

It matters a lot.

You think the PS5 would be getting mega sales if it was called the PS4 Pro Plus?

it's easier to change the name than the hardware inside the device. that's the point, the name can quickly pivot with a change in marketing direction. maybe, hypothetically speaking, Sony did originally call the PS5 the PS4 Pro Plus, but realized they'd sell more by calling it the PS5. they can make that change

 

[fwd-bwd]

To be honest I had a similar train of thought a while back, the idea that it would most likely be effective for 3rd parties who already have PC ports of games, and buying the cloud versions would also eventually translate to a digital code of the game in question once the user upgrades.

The economics of cloud gaming IMHO is still in its infancy. Most current attempts are only following the obvious (and sometime managed to screw that up still cough google cough). I at least applaud Spawn Wave for being creative here, even though his suggestion of the dual Switch 2/cloud games makes for a weak business case.

A potentially better use of the dual cloud/Switch games model is to cloudfund porting games to the Switch. Somewhat similar to the crowdfunding concept, a developer can put up a PC game on the cloud for the Switch, and promise that if enough people buy the cloud game, they will port the game natively; the buyers of the cloud game can later upgrade (sidegrade?) to the native version either for free or at a discount. So instead of port-begging, if enough Switch players actually support a game by purchasing, they essentially crowdfund cloudfund the port.

For developers, this would greatly mitigate the ROI risk of developing for the Switch. For gamers, not only they'd have a direct impact on whether a title is ported, but also be able to start playing on the Switch right away, sans the long wait or risk associated with crowdfunding.