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

[Hermii]

It was also going to be more than $400 once we knew it had 12 SM's.

Yea but nobody betted on 2024.

 

[Shoulder]

The hunger for a beefier Switch has been fierce and I remain as astonished as ever as to why nearly 7 years into Switch they still won't even begin the conversation about a successor.

Perhaps we're thinking of this all in the wrong fashion. To us, it's been nearly 7 years since the Switch launched, and we've heard absolute zip from Nintendo about it. But to Nintendo, they're riding high on the success of the system that continues to sell like crazy. We have repeat customers who bought V1 Switches that have now upgraded to OLED variants. And some have even resorted to keeping their V1 Switch's as docked only, while buying a Lite for on-the-go. Now, that latter scenario is I think rare, but I'm struggling to find a time when there are potentially millions of gamers buying a 2nd copy if you will of a console where the reasoning isn't "It broke, so I bought a new one."

It's as though Nintendo are to the Switch as to what the Toyota is to the 4Runner: Damn thing just keeps on selling, and doesn't really get many updates, let alone a full generational upgrade.

If anything, we may be astonished, but I think Nintendo are just as astonished as we are, and as a result, they'll control the narrative, and announce the new system when they want to. What I'm trying to say is if we want Nintendo to announce it sooner, then consumers better stop buying the Switch. lol

 

[Tentacle-tropes]

June is when Sizzle Season 2024 is happening, the last update before we get all 2 years of promised support.


Sales have only really been exceptional in Japan, not so much overseas. If that's because of Japan overperforming or the rest of the world underperforming is up for debate, as sequels on the same system usually sell less than the first new entry on the platform.

If you look at the sales of Splatoon 3 it is doing fine outside of JP. Launched align outside sales are doing better. The split is pretty in-line with what we saw with the other 2 games. The only reason we hear about JP more is because they value the brand more & are very vocal about it.

 

[Kevin]

Perhaps we're thinking of this all in the wrong fashion. To us, it's been nearly 7 years since the Switch launched, and we've heard absolute zip from Nintendo about it. But to Nintendo, they're riding high on the success of the system that continues to sell like crazy. We have repeat customers who bought V1 Switches that have now upgraded to OLED variants. And some have even resorted to keeping their V1 Switch's as docked only, while buying a Lite for on-the-go. Now, that latter scenario is I think rare, but I'm struggling to find a time when there are potentially millions of gamers buying a 2nd copy if you will of a console where the reasoning isn't "It broke, so I bought a new one."

It's as though Nintendo are to the Switch as to what the Toyota is to the 4Runner: Damn thing just keeps on selling, and doesn't really get many updates, let alone a full generational upgrade.

If anything, we may be astonished, but I think Nintendo are just as astonished as we are, and as a result, they'll control the narrative, and announce the new system when they want to. What I'm trying to say is if we want Nintendo to announce it sooner, then consumers better stop buying the Switch. lol

I saw a sales chart from the Nintendo investor meeting showing a significant decline in sales. The system did very very well but I think more or less most folks who wanted one, bought one. Now sales are slow in terms of hardware sales. I'm more curious about software attach rate but I don't think we have those. Still the Switch is slowing down significantly. Software probably still sells well which may be the reason Nintendo is dragging their feet but I do wonder how long they can go before people start to sell their systems and move on. I still remember how fast people abandoned the Wii and well we don't talk about Wii U lol.

 

[Neoxon]

If you look at the sales of Splatoon 3 it is doing fine outside of JP. Launched align outside sales are doing better. The split is pretty in-line with what we saw with the other 2 games. The only reason we hear about JP more is because they value the brand more & are very vocal about it.

It started strong for sure, but sales fell off in the West not too long after launch. The issue is that Nintendo hasn’t disclosed new numbers for Splatoon 3’s sales, which would give a more clear picture.

 

[Goodtwin]

It was also going to be more than $400 once we knew it had 12 SM's.

Im still betting on $399. The fact that Nintendo never dropped the price of the Switch skews the perception of what they can offer at a given price. The reason Nintendo never dropped the price of the Switch is not because they couldn't, but because they never needed to. If we were sitting here looking at the OG Switch priced at $249 and we asked how much better hardware we can get for an extra $150 in 2024, I think a T239 is within that budget. $10 worth of internal memory gets you a lot more capacity in 2024 than it did in 2017. $25 gets you a bigger higher resolution screen than it did in 2017. $40 gets you more RAM capacity in 2024 than it did 2017. Yes I am using fictitious prices, but you get my point.

 

[NineTailSage]

I would, as I've said before, pushing it that far creates compounding complexities. A completely reconceived dock, a new AC adapter in a whole new USB PD class, a cooling system that has to support 25W... When the weight of that system will be felt when it's very much not pushing that.

It's an awful lot of added complexity for what appears to be marginal gains; at 4N, I believe 25W would be well into diminishing returns.

And again.

25W cooling systems. We've seen them. For three-year-olds' hands they are not.

Nintendo handhelds, however, are.

Marginal to modest performance gains at the cost of... Appealing to children and commuters and a literal ballooning of BOM thanks to the compounding complexity and inability to reuse ANYTHING from Switch.

That's an awful big price to pay when the CPU isn't going to be keeping up anyway.

This is the company that reused the AC adapter from a WiFi network adapter up through New 2DS XL. I would be extremely surprised if they changed AC adapter, when the reason for doing so (increased power consumption) would also result in considerably higher costs and LOWER appeal.

If anything, T239 at 4N appears specifically designed NOT to exceed the power consumption parameters of the original Switch. Which makes sense, for the reasons above.

I thought this was reported some time ago especially pertaining to EU laws of why Switch 2 will most likely see a revamp on power delivery and could support higher output when docked.
Some quotes from the article...

Under the new rules, consumers will no longer need a different charger every time they purchase a new device, as they will be able to use one single charger for a whole range of small and medium-sized portable electronic devices.

Regardless of their manufacturer, all new mobile phones, tablets, digital cameras, headphones and headsets, handheld videogame consoles and portable speakers, e-readers, keyboards, mice, portable navigation systems, earbuds and laptops that are rechargeable via a wired cable, operating with a power delivery of up to 100 Watts, will have to be equipped with a USB Type-C port.

All devices that support fast charging will now have the same charging speed, allowing users to charge their devices at the same speed with any compatible charger.

Long-awaited common charger for mobile devices will be a reality in 2024 | News | European Parliament

Following Parliament’s approval, EU consumers will soon be able to use a single charging solution for their electronic devices.

www.europarl.europa.eu

 

[Shoulder]

I saw a sales chart from the Nintendo investor meeting showing a significant decline in sales. The system did very very well but I think more or less most folks who wanted one, bought one. Now sales are slow in terms of hardware sales. I'm more curious about software attach rate but I don't think we have those. Still the Switch is slowing down significantly. Software probably still sells well which may be the reason Nintendo is dragging their feet but I do wonder how long they can go before people start to sell their systems and move on. I still remember how fast people abandoned the Wii and well we don't talk about Wii U lol.

I think most dropped off from Wii not because it's Nintendo, but because it was ultimately a fad that overstayed its welcome. If anything, handheld gaming is hotter than its ever been, and now with the likes of the Steam Deck, ROG Ally, Aya NEO, and more recently the Lenovo Legion Go. There's clearly a market for handheld gaming PCs, and the same I think holds true for handheld console gaming.

Quite honestly, I could almost see Microsoft shifting gears with the Series S, and "shrinking" it down to become a Xbox Series Portable. Same power as Series S, but in portable form, potentially even as a psuedo competitor to the Switch 2. Unlikely sure, but Microsoft appears to have tried everything with Xbox, so I wouldn't put it passed them to give it a go. And yes, I'm aware some serious changes would be needed to keep the system down in terms of power usage.

EDIT: Sorry for digressing, but t answer your question, if what Nate is saying that a March reveal is likely, we may not have to wait long.

 

[Hermii]

I think most dropped off from Wii not because it's Nintendo, but because it was ultimately a fad that overstayed its welcome. If anything, handheld gaming is hotter than its ever been, and now with the likes of the Steam Deck, ROG Ally, Aya NEO, and more recently the Lenovo Legion Go. There's clearly a market for handheld gaming PCs, and the same I think holds true for handheld console gaming.

Quite honestly, I could almost see Microsoft shifting gears with the Series S, and "shrinking" it down to become a Xbox Series Portable. Same power as Series S, but in portable form, potentially even as a psuedo competitor to the Switch 2. Unlikely sure, but Microsoft appears to have tried everything with Xbox, so I wouldn't put it passed them to give it a go. And yes, I'm aware some serious changes would be needed to keep the system down in terms of power usage..

I don't think it would be literally a Series S, it just needs to be good enough to be easy to downport to (with a lower res screen/ resolution target). And honestly, you put the "velocity architecture" on Steamdeck and your already most of the way there probably, looking at what the SD can do without bespoke optimization.

 

[Aether]

I don't think it would be literally a Series S, it just needs to be good enough to be easy to downport to (with a lower res screen/ resolution target). And honestly, you put the "velocity architecture" on Steamdeck and your already most of the way there probably, looking at what the SD can do without bespoke optimization.

For it to be a Xbox and not another pc? It has to be. Downporting: another platform to develop for. Managing to put an series S into the handheld format: instant optimized Xbox games.

For streaming they don't need a handheld, every phone can do that. For pc games...well, no optimization, the ease of use is lost since games need to have specific settings/profiles like they need on Steam deck and it's no "console" anymore because of that.

 

[Hermii]

For it to be a Xbox and not another pc? It has to be. Downporting: another platform to develop for. Managing to put a series S into the handheld format: instant optimized Xbox games.

I see what you mean, but there are certain things you can do on a handheld to play to its strengths, thats not compatible with literally shrinking a Series S.

You can't have gddr bandwidth on a handheld, but you can have significantly higher capacity of lpddr without breaking the bank.

You can have a lower res screen, that will look great because of higher ppi and games running native. Which will also ease the processor load.

You can get away with a lot more compromises on that small handheld screen.

This is why, while I definitely see the benefits of literally shrinking a Series S, it woudnt play very well to h the strengths of the handheld format.

 

[Aether]

I see what you mean, but there are certain things you can do on a handheld to play to its strengths, thats not compatible with literally shrinking a Series S.

You can't have gddr bandwidth on a handheld, but you can have significantly higher capacity of lpddr without breaking the bank.

You can have a lower res screen, that will look great because of higher ppi and games running native. Which will also ease the processor load.

You can get away with a lot more compromises on that small handheld screen.

This is why, while I definitely see the benefits of literally shrinking a Series S, it woudnt play very well to h the strengths of the handheld format.

Sure, which is also why I just don't see it happening. But also if it would, then Microsoft would need a hardware/software layer that on the fly translates for hardware changes since it's unrealistic to expect all Devs to update there games, or people to be hyped about a slow dripfeed of Xbox releases to get compatibility patches. Oh and the Apu is probably even with a big die shrink not made for handhelds, and the moment you use a different one compatibility becomes impossible outside of just good old porting.

 

[Dekuman]

Even by those numbers, we aren't going to see such a game on Switch 2 hit that size like it does on PS5, simply because the device is not going to have that power to produce that graphical fidelity.

This is not how things work. Switch 2 will get current gen games. Texture quality and encoded videos may be lower quality, but it can't make a 100GB games into sub 10GB or even 20GB, very likely sitll looking at 50GB+ games on the high end, 20-25GB as common.

It will certainly be powerful enough to run current gen games.
If you're suggesting it just won't get those games at all, then I have to ask you what exactly you expect Switch 2 to be given everything else known about it.

 

[ItWasMeantToBe19]

Storage costs: $30-$50 ($30 for 256, $50 for 512)
Joycon costs: $30-$50 (Gets more expensive the more ambitious it is)
APU costs: ???
RAM costs: $20-40 (a phone analysis had 12 GBs of LPDDR5 at $20 but that seems too low...)
R&D expensing: ???
Packaging/shipping costs: ???
Marketing costs: ???
Casing: ???
Battery+dock+cords+heatpipe+wifi chip+bluetooth chip+warranty costs+outlet+card slot+power supply+etc: ???
Screen: ???

Not sure how to estimate the rest of these

 

[Discostew]

Amusingly enough, they could use it as an AI upscaler https://blogs.nvidia.com/blog/what-is-ai-upscaling/ as well as using it for any other jobs before sending the image to the screen like adding OS overlays. I don't know the pros and cons of doing that rather than relying on the main GPU though.

Unless I'm mistaken, it's not something used in gaming because of the amount of latency it introduces, which would be significantly more than what DLSS would return.

 

[Discostew]

This is not how things work. Switch 2 will get current gen games. Texture quality and encoded videos may be lower quality, but it can't make a 100GB games into sub 10GB or even 20GB, very likely sitll looking at 50GB+ games on the high end, 20-25GB as common.

It will certainly be powerful enough to run current gen games.
If you're suggesting it just won't get those games at all, then I have to ask you what exactly you expect Switch 2 to be given everything else known about it.

I wasn't suggesting it would be in the realm of 1/5 the size. Just that it would simply be less. Closer to the game size on Series S, for instance.

 

[Tentacle-tropes]

It started strong for sure, but sales fell off in the West not too long after launch. The issue is that Nintendo hasn’t disclosed new numbers for Splatoon 3’s sales, which would give a more clear picture.

I just had this conversation on IB & from what I’ve been told it is tracking higher launch aligned. Even still I doubt JP accounts for all new sales nor is disproportionately higher in the splits. I suspect when all is said & done WW will be higher like the other 2 games.

I saw a sales chart from the Nintendo investor meeting showing a significant decline in sales. The system did very very well but I think more or less most folks who wanted one, bought one. Now sales are slow in terms of hardware sales. I'm more curious about software attach rate but I don't think we have those. Still the Switch is slowing down significantly. Software probably still sells well which may be the reason Nintendo is dragging their feet but I do wonder how long they can go before people start to sell their systems and move on. I still remember how fast people abandoned the Wii and well we don't talk about Wii U lol.

They are on track to hit the 15mil guidance they set for the FY with no current revision. Software wise they increased the number. Considering software sales remain high is encouraging since it means the base is active. Considering some other information from their latest financial results we can see that people aren’t really abandoning the system.
However, sales aren’t really the reason they are “dragging” their heels; it has more to do with marketing. There is little reason to talk about it in the time between now & the FY end.

 

[BreakAtmo]

Hahah, I very much expect it to be 15W peak in TV mode. You can only add so much power before the cooling requirements become definitively un-handheld. It still needs to be small and light enough for children to pick up and play Kirby and Pokémon. Plus a lower maximum could theoretically mean better yields, since you're not pushing T239 to its absolute limits, units don't need to be able to sustain absolute peak clocks. I believe this was also done with Nintendo Switch, it didn't need the top bin of Tegra X1.

To be clear I'm talking about a PS5-style setup, where the clocks are dynamic but the power is fixed. This allows for higher gameplay clocks since they don't have to take extreme spikes into account.

My understanding is that the datamined screen size means we're already guaranteed a larger tablet that could hold a wider fan, capable of expelling more air (the Steam Deck OLED has just done that exact upgrade, except it's used for quieter cooling instead). If they were willing to spend a bit more $$$, a mobile vapor chamber would also be an option, which would improve over the current heat sink.

This, and the faster you go above a certain treshold, memory bandwidth becomes more and more of a bottleneck.

For the shader cores, but AFAIK not for the RT cores and tensor cores. If the system is going to be using the "concurrent DLSS" concept where the upscaling of the previous frame is handled alongside the shader and RT rendering of the current frame, that makes me wonder how much power the tensor cores might be taking when used concurrently. A 25W docked mode would allow for the tensor cores to be freely utilised alongside the other cores with no real cost, since the power they might be taking wouldn't be anything the shader cores could effectively use, due to the exact bandwidth limitations you mentioned. This would make 2160p DLSS with Ray Reconstruction at 60fps far more viable, only needing to be completed within 16.6ms.

 

[TLZ]

Not directly Nintendo related, but putting in hide tags because it is a hot, and unconfirmed rumor
* Hidden text: cannot be quoted. *

Nintendoomed.

 

[Deleted member 887]

Yes...and no. 1 game sold is a minimum, so it was effectively a win. How many switches are realistically only ever seeing 1 game?

I think there is ample room for differences in opinion on what the best Nintendo strategy is. But a couple of things worth noting.

Nintendo has sold 8.3 Switch "software units" per Switch sold. Software units includes DLC. Realistically speaking, yes, the average user does only buy about one game a year. I imagine you own more than 8 Nintendo games.

Increasing the BOM for a device without increasing cost doesn't increase the number of games sold for a profit, it increases the length of time that before the device becomes, itself, profitable as parts prices drop. Nintendo has, to our best guess, gone with the N4 process node. This removes the primary driver of cost reduction in the other consoles.

Them going for 2 to still make (realistically) a plus while building the Playerbase fast would be a short-term reduction in gain (still a considerable win through game sales and still available switch 1 sales), and aiming for it selling at cost in its second year.

Why would selling the hardware at a loss increase the playerbase? Let's look honestly at why Sony and Microsoft sell their hardware at a loss.

The combined sales of the PS3+360 were 173 million. The combined sales of the PS4+Xbone were 169 million - but the balance is very different. They're competing over the same pool of users.
From it's 1 million+ sellers, the PS3+360 sold ~150 million first party games. The PS4+Xbone sold ~135 million first party games. Those users are buying the same number of games.

But the 360 sold 200 million 3rd party units compared to the PS3 getting 85 million, despite being basically equal install bases. These games represent zero risk profit for the platform holder,who has no development investment.

Sony and Microsoft are subsidizing their consoles, because they're competing to be the best platform to play Grand Theft Battle Royale Soulslike. And a significant portion of their customer base is making the decision based on that. The pool of these customers is limited, it is a zero sum game.

Looking at the million plus sellers alone, Nintendo sold more first party units, than the last generation sold in total, third and first party combined. Nintendo sold 29 million units of 3rd party titles (in the million+ sellers category). Xbox One, a system that sold less than half the units, that sold only 40 million units in the same category total... sold 22 million units of 1st party software.

Nintendo's current hardware is the most compelling 3rd party platform they've made since the N64, and it still makes them the bottom player in that space. Nintendo can't push more units off the shelf by offering more storage for CoD.

Nintendo can only push more units by being a compelling way to play Nintendo games. And while I do think that Nintendo will choose to operate at a short term loss, the reasons for doing so are not the same as the other two console makers. If the average player buys 8 games over the course, then Nintendo give up the profit of 25% of their software revenue to recoup costs from the hardware. That is a huge increase in sales that are necessary to make that move worth it, with no 3rd party windfall likely to come.

That price is for me to buy the chip, in single quantity.
Nintendo's prices would be way lower, as would be the difference.
I know, every buck counts double in scale, but this has to be counterbalanced with the benefit of digital sales.

I agree there may be value in choosing to spend that money, but considering the level of competition in the space, I don't think that there is a huge margin on those devices. Nintendo's price might be half what consumers pay, but that's still a considerable cost increase.

I get trying to keep price low with "premium" components.
But I don't see 256 as premium, I see a solid base storage as a core strategic aspect in enticing digital. Additionally I'm not sure that there is a really feasible upgrade path

I 100% agree. I'm not saying that 256GB is "premium" in any global sense. I'm saying it is priced "at a premium" beyond the levels of the original Switch. Unlike, say, the 12GB of RAM, which likely are roughly the same cost now as the 4GB of the previous device.

My point wasn't that Nintendo wouldn't give us 256GB, or that 256GB was a "premium" option*. I'm saying something has to give with costs, and some aspects of your console are effectively permanent decisions for the length of the generation, and some are not. SOC is, RAM is, storage speed is. Storage quantity is not, screen quality is not, and unfortunately build quality is not. Nintendo probably has a maximum price ($400) and a maximum BOM (also $400), so we ain't gonna get everything we want.

In your case, if what you want is more storage... you're probably going to get it! It's the OLED OR DIE people who are pissed off - the first reports of an LCD were directly connected to saving costs relative to large storage.

though it was on the Steam Deck till *checks watch 4 and a half hours ago.

 

[ziggyrivers]

Maybe Nintendo constantly hammering the smooth transition and Nintendo account bit is because they're gonna sell it at a loss and compensate with software sales?

 

[Deleted member 887]

For the shader cores, but AFAIK not for the RT cores and tensor cores.

Yeah, but they can't be clocked independently. Their power is that the live inside the same structures as the shader cores, and while they can execute concurrently, they share resources (register files, cache, schedulers, bus) and that requires synced clocks. It might be possible to clock the tensor cores separately, but the design of the consumer GPUs doesn't do that, it's just the one clock domain across the GPU.

If the system is going to be using the "concurrent DLSS" concept where the upscaling of the previous frame is handled alongside the shader and RT rendering of the current frame, that makes me wonder how much power the tensor cores might be taking when used concurrently. A 25W docked mode would allow for the tensor cores to be freely utilised alongside the other cores with no real cost, since the power they might be taking wouldn't be anything the shader cores could effectively use, due to the exact bandwidth limitations you mentioned. This would make 2160p DLSS with Ray Reconstruction at 60fps far more viable, only needing to be completed within 16.6ms.

DLSS needs less bandwidth than native rendering, but it still needs access to memory - they need to read the original frame, various buffers containing additional data about the frame, the accumulated data of previous frames, and then it needs to write out the final, upscaled frame. All of this uses the memory bus.

The bandwidth limitations folks are talking about are based on the performance of Ampere hardware in desktops. It is likely that those cards could use extra bandwidth if it was available.

And if the concurrent DLSS trick is not common on PC then it likely counts as extra bandwidth load competing with the shaders

 

[carbvan]

Maybe Nintendo constantly hammering the smooth transition and Nintendo account bit is because they're gonna sell it at a loss and compensate with software sales?

I think this is possible but there'd be some fine print that goes along with it. Nintendo wouldn't sell this thing at a loss akin to MS and Sony - they lose some serious cash when selling their systems, at least at launch. I don't see Nintendo taking $100 losses easily at launch, they'd break even or barely lose anything. And to make up for that lost cash, I think they'd sell upgrade patches for Switch 1 games.

 

[ziggyrivers]

And to make up for that lost cash, I think they'd sell upgrade patches for Switch 1 games.

I was gonna suggest that: they're gonna sell those next-gen upgrade patches.

Either that, or tie next-gen upgrade patches to a more expensive tier of the Switch Online

 

[Concernt]

I was gonna suggest that: they're gonna sell those next-gen upgrade patches.

Either that, or tie next-gen upgrade patches to a more expensive tier of the Switch Online

That just seems extraordinarily silly. Next gen enhancements push CONSOLES, it's a selling point of the CONSOLE. Charging people to use power they already paid for with software they already paid for using work you've already done to help bolster the system goes down POORLY. With the main business appeal of next gen patches being MARKET APPEAL, doing something that makes it go down slowly is borderline market suicide.

Last time Nintendo did something even remotely comparable was the Wii to Wii U "upgrade fee" for Virtual Console games- a system that not only failed, but helped precipitate the cessation of Virtual Console altogether.

And thing graphical features to a subscription, man alive, what a terrible idea. Confusion during a console transition is not a smooth transition. Tying PATCHES to SUBSCRIPTIONS would be confusing for most consumers and viewed as disgusting by those that understand the situation.

If the NG Switch is marketed as "plays your Switch games but better" as well as its own games, then upgrade fees will BADLY sour consumer sentiment. The exact opposite of what you want at the launch of a console.

They're not going to sell the next gen patches. The supposed patch type for them, datapatch, doesn't even appear to be attached to the DRM, licence or DLC related elements of the system.

Technically inferior and monumentally unpopular, no enhancements at all would be far more likely than PAID.

 

[ziggyrivers]

That just seems extraordinarily silly. Next gen enhancements push CONSOLES, it's a selling point of the CONSOLE. Charging people to use power they already paid for with software they already paid for using work you've already done to help bolster the system goes down POORLY. With the main business appeal of next gen patches being MARKET APPEAL, doing something that makes it go down slowly is borderline market suicide.

Last time Nintendo did something even remotely comparable was the Wii to Wii U "upgrade fee" for Virtual Console games- a system that not only failed, but helped precipitate the cessation of Virtual Console altogether.

And thing graphical features to a subscription, man alive, what a terrible idea. Confusion during a console transition is not a smooth transition. Tying PATCHES to SUBSCRIPTIONS would be confusing for most consumers and viewed as disgusting by those that understand the situation.

If the NG Switch is marketed as "plays your Switch games but better" as well as its own games, then upgrade fees will BADLY sour consumer sentiment. The exact opposite of what you want at the launch of a console.

They're not going to sell the next gen patches. The supposed patch type for them, datapatch, doesn't even appear to be attached to the DRM, licence or DLC related elements of the system.

Technically inferior and monumentally unpopular, no enhancements at all would be far more likely than PAID.

All your points are valid and solid, I concur. The best thing Nintendo can do with a message is: this is the Switch, but way better. Your library can carry over, but it'll look and run better. Go buy the better Switch

 

[Concernt]

All your points are valid and solid, I concur. The best thing Nintendo can do with a message is: this is the Switch, but way better. Your library can carry over, but it'll look and run better. Go buy the better Switch

Exactly. The way I see it, next gen patches are a way to help rapid adoption. They don't have to touch up every title, just the ones people care about, the ones they might still play. Promising "plays your existing games better" is a fantastic way to fuel early uptake, as we saw with the current generation of home consoles. Rapid adoption should lead to increased volume of software sales; not only of next gen software, but of all software available on the platform as hardware launches spur software sales. It's more appealing and could, in the long term, also be more profitable than fee-to-play upgrades, just indirectly. I think of it like one of their "value added services", like the NSO app or captures. It doesn't directly increase revenue, but increases consumer engagement, which drives software sales.

 

[ziggyrivers]

They don't have to touch up every title,

I feel like Nintendo's Evergreens will be considered for this. Hell, can we expect monthly updates of Evergreens being patched? Like a sort of Software as a Service, but for the sake of pacing the idea of next gen patches. Would that be too rough?

 

[Discostew]

Last time Nintendo did something even remotely comparable was the Wii to Wii U "upgrade fee" for Virtual Console games- a system that not only failed, but helped precipitate the cessation of Virtual Console altogether.

Personally, my thoughts on the Wii U "upgrade" of VC was less about Nintendo trying to get more money, but was for dealing with 3rd-party licensing. The transfer from Wii to Wii U was free, but the games were still stuck in the Wii Mode environment. When getting the upgrade, you got a completely new copy of the game at a discount, not a patch. You retained the original Wii VC, but that was more or less redundant because of copy accessible from the Wii U menu. Nintendo could have made their own VC games free of charge from Wii Mode to Wii U, but that may have conflicted with 3rd-parties.

 

[boomersooner2000]

Storage costs: $30-$50 ($30 for 256, $50 for 512)
Joycon costs: $30-$50 (Gets more expensive the more ambitious it is)
APU costs: ???
RAM costs: $20-40 (a phone analysis had 12 GBs of LPDDR5 at $20 but that seems too low...)
R&D expensing: ???
Packaging/shipping costs: ???
Marketing costs: ???
Casing: ???
Battery+dock+cords+heatpipe+wifi chip+bluetooth chip+warranty costs+outlet+card slot+power supply+etc: ???
Screen: ???

Not sure how to estimate the rest of these

Why are Joycons so expensive to make btw? Are they more expensive to produce in comparison to PS5's dual sense controller?

 

[ItWasMeantToBe19]

Why are Joycons so expensive to make btw? Are they more expensive to produce in comparison to PS5's dual sense controller?

I have no idea how much the dual sense costs to make. Would love an estimate if one existed.

The PS4 controller cost around $18 to make in 2013 and the joycons have a lot more going on. I expect the next gen joycons to be way more advanced than the current gen ones as well. If there is a unique gimmick, it has to be on the joycons as well and that could raise prices too.

 

[Concernt]

I feel like Nintendo's Evergreens will be considered for this. Hell, can we expect monthly updates of Evergreens being patched? Like a sort of Software as a Service, but for the sake of pacing the idea of next gen patches. Would that be too rough?

That wouldn't be out of the ordinary with what we saw for Xbox Series X, with some at launch, and several coming bit by bit at a later date. I sort of expect an "as they're ready" scenario. Some, like BOTW and Splatoon 3 ready on launch day. Then we might get Bayonetta 3 a few months later, and so on. It's a way to re-up interest in their evergreens and increase engagement with their new console during possible lulls in new software releases.

As I've said before, as optimistic as I am about the hardware and its launch window, I am a pessimist about software. I expect a weak first year, really. Super Mario Galaxy 2 4K, the HD Zeldas, patches for Switch games, BDSP style Black and White remakes, with the only real headliners I could think of actually managing to hit launch year being Prime 4 and 3D Mario. If I were to bet, Prime 4 at launch and 3D Mario for the holidays. Unlike Nintendo Switch this isn't a complete reset on the library or the brand, so they don't have to have that carefully constructed, ever so lucky, perfectly planned first year Switch has. As long as there's something worth playing in the first year, and enhancements for Switch games (even if those are essentially new versions under the hood), it should have no problem achieving year one sales goals. People are hungry for new Nintendo hardware. Feed them first, and release software when it's ready.

That said, I would be very happy to be wrong about that and see an absurd year of mega-sized Animal Crossing, ARMS2, Mario Kart 10 and so forth. I just don't expect it.

 

[BreakAtmo]

Yeah, but they can't be clocked independently. Their power is that the live inside the same structures as the shader cores, and while they can execute concurrently, they share resources (register files, cache, schedulers, bus) and that requires synced clocks. It might be possible to clock the tensor cores separately, but the design of the consumer GPUs doesn't do that, it's just the one clock domain across the GPU.

Oh, I wasn't suggesting otherwise. I know the whole chip is clocked the same across all cores. But in a TDP-starved environment, isn't using shader+RT+tensor cores continuously and simultaneously going to use more power at once than shader+RT cores and tensor cores separately?

DLSS needs less bandwidth than native rendering, but it still needs access to memory - they need to read the original frame, various buffers containing additional data about the frame, the accumulated data of previous frames, and then it needs to write out the final, upscaled frame. All of this uses the memory bus.

That's fair, but would it use more or less than the shader cores being pushed to super-high clocks?

The bandwidth limitations folks are talking about are based on the performance of Ampere hardware in desktops. It is likely that those cards could use extra bandwidth if it was available.

And if the concurrent DLSS trick is not common on PC then it likely counts as extra bandwidth load competing with the shaders

My assumption is that concurrent DLSS isn't used on PC because the desktop cards almost never need it. When even the weaker cards (that probably wouldn't target 2160p anyway) take only about 3ms to make a 2160p DLSS frame, it's not worth it to bother programming in the concurrency option, because the extra frame of latency probably won't be worth it. The Switch 2 is where the concept will actually be worthwhile to use.

I understand that the technical considerations are more complex than I thought, but in the big picture, do you think a fixed 25W power draw in docked mode would improve things enough to be worth it?

 

[redmutineer75]

I think there is ample room for differences in opinion on what the best Nintendo strategy is. But a couple of things worth noting.

Nintendo has sold 8.3 Switch "software units" per Switch sold. Software units includes DLC. Realistically speaking, yes, the average user does only buy about one game a year. I imagine you own more than 8 Nintendo games.

Increasing the BOM for a device without increasing cost doesn't increase the number of games sold for a profit, it increases the length of time that before the device becomes, itself, profitable as parts prices drop. Nintendo has, to our best guess, gone with the N4 process node. This removes the primary driver of cost reduction in the other consoles.


Why would selling the hardware at a loss increase the playerbase? Let's look honestly at why Sony and Microsoft sell their hardware at a loss.

The combined sales of the PS3+360 were 173 million. The combined sales of the PS4+Xbone were 169 million - but the balance is very different. They're competing over the same pool of users.
From it's 1 million+ sellers, the PS3+360 sold ~150 million first party games. The PS4+Xbone sold ~135 million first party games. Those users are buying the same number of games.

But the 360 sold 200 million 3rd party units compared to the PS3 getting 85 million, despite being basically equal install bases. These games represent zero risk profit for the platform holder,who has no development investment.

Sony and Microsoft are subsidizing their consoles, because they're competing to be the best platform to play Grand Theft Battle Royale Soulslike. And a significant portion of their customer base is making the decision based on that. The pool of these customers is limited, it is a zero sum game.

Looking at the million plus sellers alone, Nintendo sold more first party units, than the last generation sold in total, third and first party combined. Nintendo sold 29 million units of 3rd party titles (in the million+ sellers category). Xbox One, a system that sold less than half the units, that sold only 40 million units in the same category total... sold 22 million units of 1st party software.

Nintendo's current hardware is the most compelling 3rd party platform they've made since the N64, and it still makes them the bottom player in that space. Nintendo can't push more units off the shelf by offering more storage for CoD.

Nintendo can only push more units by being a compelling way to play Nintendo games. And while I do think that Nintendo will choose to operate at a short term loss, the reasons for doing so are not the same as the other two console makers. If the average player buys 8 games over the course, then Nintendo give up the profit of 25% of their software revenue to recoup costs from the hardware. That is a huge increase in sales that are necessary to make that move worth it, with no 3rd party windfall likely to come.




I agree there may be value in choosing to spend that money, but considering the level of competition in the space, I don't think that there is a huge margin on those devices. Nintendo's price might be half what consumers pay, but that's still a considerable cost increase.


I 100% agree. I'm not saying that 256GB is "premium" in any global sense. I'm saying it is priced "at a premium" beyond the levels of the original Switch. Unlike, say, the 12GB of RAM, which likely are roughly the same cost now as the 4GB of the previous device.

My point wasn't that Nintendo wouldn't give us 256GB, or that 256GB was a "premium" option*. I'm saying something has to give with costs, and some aspects of your console are effectively permanent decisions for the length of the generation, and some are not. SOC is, RAM is, storage speed is. Storage quantity is not, screen quality is not, and unfortunately build quality is not. Nintendo probably has a maximum price ($400) and a maximum BOM (also $400), so we ain't gonna get everything we want.

In your case, if what you want is more storage... you're probably going to get it! It's the OLED OR DIE people who are pissed off - the first reports of an LCD were directly connected to saving costs relative to large storage.

though it was on the Steam Deck till *checks watch 4 and a half hours ago.

In that regard, storage is likely much higher on the priority list than the screen. If they want big third-party titles like the next Call of Duty, Final Fantasy, or Monster Hunter, they're gonna have to up that file size, and the people who are buying graphically intensive games are likely playing them docked for the most performance, and thus the screen is irrelevant to them. Sure, people will talk up how LCD is a downgrade, but at the end of the day, having more games available is more important to Nintendo than those games looking nicer portable.

 

[redmutineer75]

As I've said before, as optimistic as I am about the hardware and its launch window, I am a pessimist about software. I expect a weak first year, really. Super Mario Galaxy 2 4K, the HD Zeldas, patches for Switch games, BDSP style Black and White remakes, with the only real headliners I could think of actually managing to hit launch year being Prime 4 and 3D Mario.

That assumes that they haven't been working on new titles for it that we just don't know about. We tend to forget that leaks are the exception, not the rule.

 

[Concernt]

That assumes that they haven't been working on new titles for it that we just don't know about. We tend to forget that leaks are the exception, not the rule.

I try not to forget that! We don't really have much evidence to either side of "lots of" or "little" software- but I remain pessimistic. They've undoubtedly been working on things. I just don't expect them to hit launch year!

Some of this is just how I personally feel and has no real basis- I'm a software pessimist. Some of this is just how recent major titles are. Splatoon, Zelda, Animal Crossing, Pokémon, 2D Mario, Pikmin, etc. all within the last 2-3 years- so unfortunately those are likely out of the running for new games in the first year of the next system, and those are some big names.

3D Mario, Metroid, and Mario Kart seem the most likely. Of those, 3D Mario had new content in 2021, and Mario Kart 8 Deluxe only just finished up DLC.

I look forward to the hardware, but the software side doesn't seem promising. I do hope I'm wrong, though!

I look forward to the hardware anyway. I think the big appeal in the first year will be enhancements for Switch games and big third party titles brought over to Nintendo for the first time.

 

[Neoxon]

That wouldn't be out of the ordinary with what we saw for Xbox Series X, with some at launch, and several coming bit by bit at a later date. I sort of expect an "as they're ready" scenario. Some, like BOTW and Splatoon 3 ready on launch day. Then we might get Bayonetta 3 a few months later, and so on. It's a way to re-up interest in their evergreens and increase engagement with their new console during possible lulls in new software releases.

As I've said before, as optimistic as I am about the hardware and its launch window, I am a pessimist about software. I expect a weak first year, really. Super Mario Galaxy 2 4K, the HD Zeldas, patches for Switch games, BDSP style Black and White remakes, with the only real headliners I could think of actually managing to hit launch year being Prime 4 and 3D Mario. If I were to bet, Prime 4 at launch and 3D Mario for the holidays. Unlike Nintendo Switch this isn't a complete reset on the library or the brand, so they don't have to have that carefully constructed, ever so lucky, perfectly planned first year Switch has. As long as there's something worth playing in the first year, and enhancements for Switch games (even if those are essentially new versions under the hood), it should have no problem achieving year one sales goals. People are hungry for new Nintendo hardware. Feed them first, and release software when it's ready.

That said, I would be very happy to be wrong about that and see an absurd year of mega-sized Animal Crossing, ARMS2, Mario Kart 10 and so forth. I just don't expect it.

If they don’t have the software ready, it won’t launch. Having big game after big games spearheaded by a heavy-hitter on Day 1 helped the Switch 1 be a massive hit from the jump. Launching with a game that an entire region doesn’t care for with barely anything substantial until the holidays is a recipe for disaster, especially since Japan was instrumental in the Switch 1’s success. And enhanced older games alone won’t cut it.

 

[Tof]

I reckon that we might get a reveal trailer in Jan and what Nate heard about March is about the presentation (similar to Jan 2017). This makes more sense to me than a reveal trailer in March because a lot of developers would want working builds of their games running on the new console so that it can be played live on stage whereas in a trailer, footage from elsewhere can be and usually is just edited into the trailer and isn’t running on the actual console.
I think we might then see a release in May (or June). Not sure how good shareholder Chad sources were, but it seems at least some investors with inside knowledge (which has the same dubious legality as a leak) have acted according to a May release.

This is my head canon crackpot assumption as of right now but if there are any holes or other factors to consider then please let me know and I can adjust my theory accordingly.

 

[Samuspet]

If nintendo wants games like Baldur's Gate 3 and GTA 6 to have switch 2 version, having only 128GB of internal storage kills any chance of that, especially because unlike with the switch 1, using microSD for expandable storage likely won't be possible for the switch 2 due to the limited speeds of microSD.

Nintendo will likely sell 150 million Switch consoles by the end of 2024 (when I expect Switch 2 to launch) without major AAA third party support. I don’t really see that changing nor third parties view that Nintendo are still almost a full generation behind PS and Xbox so why bother with the headache with regards to file sizes and Nintendo’s anemic online service as a whole (see Kotick’s comments about Switch 2 being PS4 gen).

With AAA releases becoming less with each generation considering how long they take to create and the return of AA, the continued strength of indies and the explosion / continuation of service games like Minecraft, Fortnite, Apex aswell as there being potentially hundreds of PS4/XBO ports that never came to the original Switch which are smaller in file size, I really don’t disagree with Nintendo IF they hold the view that they don’t need to have anymore than 128GB of fast storage at launch.

 

[Concernt]

I reckon that we might get a reveal trailer in Jan and what Nate heard about March is about the presentation (similar to Jan 2017). This makes more sense to me than a reveal trailer in March because a lot of developers would want working builds of their games running on the new console so that it can be played live on stage whereas in a trailer, footage from elsewhere can be and usually is just edited into the trailer and isn’t running on the actual console.
I think we might then see a release in May (or June). Not sure how good shareholder Chad sources were, but it seems at least some investors with inside knowledge (which has the same dubious legality as a leak) have acted according to a May release.

This is my head canon crackpot assumption as of right now but if there are any holes or other factors to consider then please let me know and I can adjust my theory accordingly.

 

[Deleted member 887]

Why are Joycons so expensive to make btw? Are they more expensive to produce in comparison to PS5's dual sense controller?

Joy-Cons are pricey because they’re so damn small, and because there are two of them. Just about every part is more expensive - the BT chip, the gyroscope, the charging module all have to fit in a device small than a flip phone. And each joy con needs all the hardware, so it’s all doubled.

I understand that the technical considerations are more complex than I thought, but in the big picture, do you think a fixed 25W power draw in docked mode would improve things enough to be worth it?

No idea really. I mean, obviously it would perform better, significantly. But “worth it” is tangled all with handheld as well. Like, is pushing past the bandwidth limits add performance? Sure. Does doing it in docked affect battery life at all? Nope!

But does it mean that handheld and docked require radically different optimization strategies? Maybe. Does it require a bigger fan - probably and that might impact handheld.

To be honest I was just reacting to the clicking specifics, I don’t have a solid on the broader questions.

 

[redmutineer75]

I try not to forget that! We don't really have much evidence to either side of "lots of" or "little" software- but I remain pessimistic. They've undoubtedly been working on things. I just don't expect them to hit launch year!

Some of this is just how I personally feel and has no real basis- I'm a software pessimist. Some of this is just how recent major titles are. Splatoon, Zelda, Animal Crossing, Pokémon, 2D Mario, Pikmin, etc. all within the last 2-3 years- so unfortunately those are likely out of the running for new games in the first year of the next system, and those are some big names.

3D Mario, Metroid, and Mario Kart seem the most likely. Of those, 3D Mario had new content in 2021, and Mario Kart 8 Deluxe only just finished up DLC.

I look forward to the hardware, but the software side doesn't seem promising. I do hope I'm wrong, though!

I look forward to the hardware anyway. I think the big appeal in the first year will be enhancements for Switch games and big third party titles brought over to Nintendo for the first time.

I mean, that tends to be how it happens with Nintendo. Again, leaks are the exception, not the rule. A lot of people were insistent on a 2023 launch, but that all went kaput once Wonder was announced. Simply put, none of us are devs at Nintendo (that we know of) and can't make judgements one way or the other, but given their track record, I would be surprised if they broke away from their monthly first party release schedule.

 

[Concernt]

Again, leaks are the exception, not the rule.

...again?

I... Specifically didn't mention leaks. I think you might be reading some implication that isn't there.

I'm not basing this on leaks? I'm basing this expectation purely on the existing recent releases of major titles and my own feelings on the matter (I.E., software won't be as essential to push this device early on compared to Switch, because it isn't a library reset.).

This is the next generation of Nintendo Switch - it'll sell like hotcakes regardless. The hardware itself is sure to be appealing.

 

[BreakAtmo]

No idea really. I mean, obviously it would perform better, significantly. But “worth it” is tangled all with handheld as well. Like, is pushing past the bandwidth limits add performance? Sure. Does doing it in docked affect battery life at all? Nope!

But does it mean that handheld and docked require radically different optimization strategies? Maybe. Does it require a bigger fan - probably and that might impact handheld.

To be honest I was just reacting to the clicking specifics, I don’t have a solid on the broader questions.

Fair enough. I don't think the two modes would have to be radically different, though - the docked mode is now targeting quadruple the resolution instead of double, so I was thinking the extra power could help with that. And I'd expect concurrent DLSS to be used in both, if it is used. And I was also thinking how the bigger fan (and maybe vapor chamber) would affect the handheld mode - it would likely make it notably quieter and/or cooler, still at 7W. Leave a 15W handheld mode for the Switch 3.

Of course I also have no idea how much a mobile vapor chamber would cost. Less than a big one, I'd hope.

 

[P4bl0]

I reckon that we might get a reveal trailer in Jan and what Nate heard about March is about the presentation (similar to Jan 2017). This makes more sense to me than a reveal trailer in March because a lot of developers would want working builds of their games running on the new console so that it can be played live on stage whereas in a trailer, footage from elsewhere can be and usually is just edited into the trailer and isn’t running on the actual console.
I think we might then see a release in May (or June). Not sure how good shareholder Chad sources were, but it seems at least some investors with inside knowledge (which has the same dubious legality as a leak) have acted according to a May release.

This is my head canon crackpot assumption as of right now but if there are any holes or other factors to consider then please let me know and I can adjust my theory accordingly.

Maybe skip the trailer and presentation straight

 

[Gerald]

Storage costs: $30-$50 ($30 for 256, $50 for 512)
Joycon costs: $30-$50 (Gets more expensive the more ambitious it is)
APU costs: ???
RAM costs: $20-40 (a phone analysis had 12 GBs of LPDDR5 at $20 but that seems too low...)
R&D expensing: ???
Packaging/shipping costs: ???
Marketing costs: ???
Casing: ???
Battery+dock+cords+heatpipe+wifi chip+bluetooth chip+warranty costs+outlet+card slot+power supply+etc: ???
Screen: ???

Not sure how to estimate the rest of these

im bet 449$, costs $20 for 12GB LPPPDR5 RAM can be accurate since RAM price drop very significantly over the last year, for storage cost you check UFS 3.0/3.1?

 

[Neoxon]

...again?

I... Specifically didn't mention leaks. I think you might be reading some implication that isn't there.

I'm not basing this on leaks? I'm basing this expectation purely on the existing recent releases of major titles and my own feelings on the matter (I.E., software won't be as essential to push this device early on compared to Switch, because it isn't a library reset.).

This is the next generation of Nintendo Switch - it'll sell like hotcakes regardless. The hardware itself is sure to be appealing.

That only worked for Sony because they have third-party support fully locked down. Nintendo can’t count on that, plus it’s not what people buy Nintendo systems for anyway.

For most, Nintendo systems are a means to play Nintendo games. And Nintendo needs to have a steady stream of new games if they want the Switch 2 to be anywhere near as successful as the original.

 

[Detective Chimp]

* Hidden text: cannot be quoted. *

People seem to be treating this as a standard console launch. With months of marketing etc. I really think they’re going the smartphone route. Announcement and launch shortly after. Aggressive marketing throughout the year and leading up to holiday period for a second launch essentially. Worked wonderfully for the OG Switch. Don’t see why they would do it any different. Other than announcing it closer to launch.

 

[ItWasMeantToBe19]

im bet 449$, costs $20 for 12GB LPPPDR5 RAM can be accurate since RAM price drop very significantly over the last year, for storage cost you check UFS 3.0/3.1?

I'm looking at UFS 3.1 estimates from recent phones.

 

[Gerald]

I'm looking at UFS 3.1 estimates from recent phones.

oh nice, btw can you check estimate prices for LPDD5 16GB?