• Hey everyone, staff have documented a list of banned content and subject matter that we feel are not consistent with site values, and don't make sense to host discussion of on Famiboards. This list (and the relevant reasoning per item) is viewable here.

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

Yeah, they aim to look retro, but they've got a lot going on and run decidedly sub-native.

One doesn't have to look farther than the only two systems to sell more than 150m to find a couple good examples of that.

That only shows through last March--but not a lot has changed in the meantime. Slightly up, then slightly down.
ir02YWA.png
That graph...wow. If Nintendo wanted to release Switch 2 in spring 2025 I really think they could. Software shipments would still be at the level of the Wii after 5 years.
 
That graph...wow. If Nintendo wanted to release Switch 2 in spring 2025 I really think they could. Software shipments would still be at the level of the Wii after 5 years.
Repeating the Wii U fiasco but with an even bigger gap between consoles seems like a horrible idea.

It's not really about "how long can they drag this out" and more "when is the right tine to strike".

Consider how good GBA software was doing briefly before DS launch.
 
Some folks in the thread I posted yesterday believe there will be a bigger gap between Switch 2/PS5 than there was between Switch/PS4.

Is this based on anything concrete? I don't get it.

Not saying it is. But there are people in other threads saying there will be a wider gap between PS5 and Switch than there was between PS4 and Switch, and thus downporting will be harder. That is what I was asking about.
Who is is saying there's gonna be wide gaps?

The CPU gap should be narrower. Especially if it ends up with 8 cores (7 for gaming) and if it's a respectable 1.5Ghz per core. I would say CPU wise it would only be the same as last year or worse if we got 1Ghz CPU cores all over.

GPU is hard to say. 3-4 tflops will help a lot. 2.5 is imo the minimum I hope to get. But 3 will kind of put is between PS4 vs switch GPU gap. I don't know how much ps5/x series will be evolving in terms of the AMD version of DLSS, but DLSS should be as well (at least DLSS 2). I do think DLSS will help bridge the performance gap a little though.

I think if Drake gets like 1 GB/s UFS speed for the SSD, we'll be okay for the most part of that category.

Drake's Bandwidth will likely be 102 GB/s, but I'm really hoping we get more like lpddr5x speeds. We definitely won't be able to run PS4 quality games at more than 1080p without DLSS. Not that it matters that much.

Drake's biggest bottlenecks will definitely be CPU and then maybe RAM bandwidth as second. GPU is scalable at least. 30 fps current gen games might not be possible on Drake/Switch 2, assuming that the CPU is being pushed really hard.

It's definitely more future proof for the CPU to be higher and sacrifice the GPU to achieve that at a respectable power draw. I'd rather have a 3 tflops GPU and 2Ghz CPU cores than 4 tflops GPU and 1.5 GHz CPU. This is assuming we get a 4nm tsmc node..
 
Last edited:
That depends on what you consider "full advantage".

A PS5 game that takes "full advantage" of the console by rendering the game at 4K could well survive on Drake by rendering at 540p, 1080p DLSS, and making up for the weaker CPU by reallocating tasks to the GPU and AI portions of the SOC.

This is what happened with the Witcher 3, a game that is genuinely very well optimised but also incredibly demanding, that got onto Switch despite pushing the 8th generation consoles, by reallocating CPU tasks to the GPU and making room for them by lowering the rendering resolution.

I'm oversimplifying here but, the gap may not be small, but it's small enough that those edge cases are more a case of "it would take an extra 1-2 years to optimise this to make it run acceptably on Switch Drake" rather than it being impossible.
I meant it more as a game that was developed with the PS5 as the target spec. That game could be ported to the Series X with relative ease because of how similar the two consoles are, and downscaled further to the Series S as it's just a cut-down X. Because of architectural differences, however, devs won't be able to just 'Series S' a Switch 2 version of the game. But that was always going to be the case anyway and it's not something Nintendo can fix, short of somehow squeezing a mini-Series S into a tablet form factor.

If a dev decides not to bring over a game for this reason, closing the gap by clocking the CPU/GPU higher or adding more RAM isn't gonna change their mind.
 
How likely is it we get lpddr5x? I mean, it was announced late 2021. Is it being manufactured yet? It seems like it would be mature enough (or widely available enough) for a late 2023/early 2024 Drake launch.
 
Well, inevitably, 3rd parties will begin taking full advantage of PS5 hardware and that, some fear, might put their games out of the Switch's league entirely.
Yeah. It will happen. Just as there are PS4 games that aren't getting onto Switch. The PS5 will be more more powerful than anything you can fit in a tablet form factor, and an exclusive designed to squeeze every bit of juice out of the full complement of the system's many features - it's screamingly fast storage, its very powerful single core CPU, it's 3D sound engine - then you will wind up with a game that might be hard to port even to the Xbox Series X, a very similar machine.

I am betting the number of games that fit this bill will be approximately 0. Not only is designing a game that uses that combination of features an unlikely scenario by itself, but non-exclusives will want to run on a PC, and that means a certain amount of scalability.

What's left are simply the games that are too expensive to port. And "too expensive to port" has as much to do with "how much cash will we make back from this thing" as it does with "how much cash will it take to squeeze this thing down."

The real question is, how much of the multiplat library will come to Nintendo Next, and how many of those games will be too shitty to play? If I didn't have Drake's leaked specs to look at, I'd answer that question with "about the same as now". Since I do have Drake's specs to look at I'd say "about the same as now."

Nintendo lucked out a little with Switch. The PS4/Xbox One CPUs were just godawful, and the Xbone was underpowered, and continued to put lots of games in Switch's sights, even after the One and One S launched.

This time around, the CPU situation isn't as favorable. On the other hand, Drake is looking to pack a bigger GPU, Microsoft is setting the bar once again with the Series S, and DLSS isn't a magic bullet, but is a force multiplier.

In the end it's going to come down to how well the thing sells.

IMO, the one place where I'd be concerned about the gap would be single thread CPU strength.
Hard agree. Not just for main thread programming, but for RT. Drake has those excellent RT cores, it's an open question to me whether or not Nintendo can set those cores up for success with a constrained single core setup and 8GB of RAM.

I meant it more as a game that was developed with the PS5 as the target spec. That game could be ported to the Series X with relative ease because of how similar the two consoles are, and downscaled further to the Series S as it's just a cut-down X. Because of architectural differences, however, devs won't be able to just 'Series S' a Switch 2 version of the game. But that was always going to be the case anyway and it's not something Nintendo can fix, short of somehow squeezing a mini-Series S into a tablet form factor.

If a dev decides not to bring over a game for this reason, closing the gap by clocking the CPU/GPU higher or adding more RAM isn't gonna change their mind.
I mean, as long as they're on Nvidia/ARM they're not going to be the same architecture no matter how much power they squeeze in there. This is part of why the "X% of a Series X" talk drives me nuts. The arches are different enough across the board that it's a different set of performance tradeoffs, even if the power is roughly comparable.


How likely is it we get lpddr5x? I mean, it was announced late 2021. Is it being manufactured yet? It seems like it would be mature enough (or widely available enough) for a late 2023/early 2024 Drake launch.
The thing is that the memory controller is baked into the SOC. I expect LPDDR5, same as Orin, but @Look over there has made a reasonable argument that going with 5X might save Nintendo money in the long run - because it's baked into the SOC, going with 5X future-proofs Nintendo against the older standard getting taken over by the newer one in the mobile market.
 
I don't think there is as much of a demand, for a new Switch, as you all would like to believe.
I don’t think there is a heavy demand for a new Switch. I don’t think there is ever heavy demand for an upgraded piece of consumer electronics.

Upgrades create demand. That’s the point of new features and exclusives. But there are a lot of really hot multiplats that aren’t coming to Switch. Putting on of them on a next gen portable would restore the feeling of “playing big games on the go” that the original Switch created.

The early adopters who drive that first year of sales are looking at a Steam Deck and thinking, do I need to wait for a new Switch? And if I don’t, how long will the cross gen period be? Because if it’s gonna be a while, then I can get a Steam Deck now, and hold on to my base Switch for the next Mario…
 
0
Do you guys think they’ll ever put Nintendo Switch OLED models out there on lower price?

Currently they are the best selling SKUs, beating the OG Switch models even
 
0
Yeah software is not stable. It's going to finish down ~15% from last FY.

Original forecast for this FY was 210m units, down from 235m in the previous FY. That forecast was revised downward in the recent Q3 results to 205m.
Either they're lowballing or they expect the last quarter to be pretty awful. That would put the quarter at 33m, versus 46, 55, 56 the previous three years.
 
The point is rather that - relative to other Nintendo systems like DS and Wii - the software decline is slower from a higher peak. If you put the comparison on a graph, the difference is noticeable.

I'm not pointing that out to say new hardware won't happen; I think Nintendo have new hardware out sometime between the end of this year and the middle of next, before Switch slows any further, and I think that's the right thing to do.
Oh for sure, Switch has enjoyed an incredible mid-life and tail. COVID and chip shortages were a big boon to pushing money into their ecosystem, and holding onto it longer than normal in conjunction with the great software releases and the rise of digital sales for them. Just providing the actuals for what's going on in this FY, where we can more definitively say the decline is in full effect now that software is dipping, and doing so a bit faster than Nintendo thought.
Either they're lowballing or they expect the last quarter to be pretty awful. That would put the quarter at 33m, versus 46, 55, 56 the previous three years.
There would have been no reason to revise down if they weren't pretty sure it was going to be that bad. Seems very plausible that the last quarter is going to be ugly after what we just saw in Q3. They probably over shipped heavily in the previous quarter for holidays as per usual, and still experienced a substantial dropoff, which would probably mean there is plenty of unsold hardware and software for Q4. They have no big new software to ship like Arceus, so nothing to really drive software sales.

I don't think hitting either revised number is really guaranteed either. Could be a bit similar situation to 2018-2019, where they revised a couple times, and in the end still came in slightly under.
 
Is it feasible at this point in time to for them to mass produce hall effect sticks as a drop in replacement for joy con while keeping the rest of the design the same?
 
the next MK game has probably had the longest development cycle of any Mario Kart title ever. and I believe it will blow people away when it's eventually shown, just like MK8 did originally. there's no way this title hasn't been worked on for years already, the DLC is just throwing us a bone so they can hold it back & keep finessing things until the next system. MK8 still looks great but as far as being about as good a Mario Kart title can possibly look, no way.
 
The thing is that the memory controller is baked into the SOC. I expect LPDDR5, same as Orin, but @Look over there has made a reasonable argument that going with 5X might save Nintendo money in the long run - because it's baked into the SOC, going with 5X future-proofs Nintendo against the older standard getting taken over by the newer one in the mobile market.

Drake has had car stuff remove from it which it`s orin counter part has. To use a car metafor. When they already have opened the hold to remove these why not replace the 128bit memory controller with the 256bit that the Orin AGX has as well?
 
Last edited:
Mario Kart 9 already exists. It's called Mario Kart: Tour, and it's a live service title. MK8DX receiving its tracks shows, and it being connected to NSO, is a model for Nintendo to do live service themselves. Spend 60 bucks for a base number of tracks, pay 50 bucks a year for more tracks on a drip feed.

Not saying MKX can't or won't happen. But I have 0 confidence that we'll get another Mario Kart title anytime in the next half decade.

I'm going to a wild, crazy shot in the dark here and say, yes Nintendo will make a sequel to the biggest selling non-bundled game in their history.

I also will boldly predict that Disney will make more Marvel movies.

The reason the current Switch didn't get MK9 or X or however you want to call it is because 8 Deluxe simply sold so well that it wasn't needed. Most people didn't have a Wii U and never played the game there.

And of course, locking people in to NSO subs will be a big part of it, part of why it probably will come early in the product cycle.
 
Drake has had car stuff remove from it which it`s orin counter has. To use a car metafor. When they already have opened the hold to remove these why not replace the 128bit memory controller with the 256bit that the Orin AGX has as well?
I dunno if that's ever been done on a handheld PC or will be happening.. Besides space, cost and heat/power draw are gonna be an issue.
 
Yeah. It will happen. Just as there are PS4 games that aren't getting onto Switch. The PS5 will be more more powerful than anything you can fit in a tablet form factor, and an exclusive designed to squeeze every bit of juice out of the full complement of the system's many features - it's screamingly fast storage, its very powerful single core CPU, it's 3D sound engine - then you will wind up with a game that might be hard to port even to the Xbox Series X, a very similar machine.

I am betting the number of games that fit this bill will be approximately 0. Not only is designing a game that uses that combination of features an unlikely scenario by itself, but non-exclusives will want to run on a PC, and that means a certain amount of scalability.

What's left are simply the games that are too expensive to port. And "too expensive to port" has as much to do with "how much cash will we make back from this thing" as it does with "how much cash will it take to squeeze this thing down."

The real question is, how much of the multiplat library will come to Nintendo Next, and how many of those games will be too shitty to play? If I didn't have Drake's leaked specs to look at, I'd answer that question with "about the same as now". Since I do have Drake's specs to look at I'd say "about the same as now."

Nintendo lucked out a little with Switch. The PS4/Xbox One CPUs were just godawful, and the Xbone was underpowered, and continued to put lots of games in Switch's sights, even after the One and One S launched.

This time around, the CPU situation isn't as favorable. On the other hand, Drake is looking to pack a bigger GPU, Microsoft is setting the bar once again with the Series S, and DLSS isn't a magic bullet, but is a force multiplier.

In the end it's going to come down to how well the thing sells.


Hard agree. Not just for main thread programming, but for RT. Drake has those excellent RT cores, it's an open question to me whether or not Nintendo can set those cores up for success with a constrained single core setup and 8GB of RAM.


I mean, as long as they're on Nvidia/ARM they're not going to be the same architecture no matter how much power they squeeze in there. This is part of why the "X% of a Series X" talk drives me nuts. The arches are different enough across the board that it's a different set of performance tradeoffs, even if the power is roughly comparable.



The thing is that the memory controller is baked into the SOC. I expect LPDDR5, same as Orin, but @Look over there has made a reasonable argument that going with 5X might save Nintendo money in the long run - because it's baked into the SOC, going with 5X future-proofs Nintendo against the older standard getting taken over by the newer one in the mobile market.

This makes me concerned for Switch 2/Drake's future.

But I guess we'll just have to wait and see.
 
the next MK game has probably had the longest development cycle of any Mario Kart title ever. and I believe it will blow people away when it's eventually shown, just like MK8 did originally. there's no way this title hasn't been worked on for years already, the DLC is just throwing us a bone so they can hold it back & keep finessing things until the next system. MK8 still looks great but as far as being about as good a Mario Kart title can possibly look, no way.
I just don't know what the next Mario Kart could possibly do to separate itself from MK8D. I mean we're going to have 70 tracks in MK8D, and now we're getting another 4 new characters.
 
the next MK game has probably had the longest development cycle of any Mario Kart title ever. and I believe it will blow people away when it's eventually shown, just like MK8 did originally. there's no way this title hasn't been worked on for years already, the DLC is just throwing us a bone so they can hold it back & keep finessing things until the next system. MK8 still looks great but as far as being about as good a Mario Kart title can possibly look, no way.
Depends, really. Remember that same team made ARMS and worked on (as in assisted but didn't spearhead) Tour and the BCP.

I wouldn't think Mario Kart 10 or X or 9 or whatever has had nearly as much time in the oven as some people think. It certainly hasn't had 10. Closer to 5 or 6.
 
0
I just don't know what the next Mario Kart could possibly do to separate itself from MK8D. I mean we're going to have 70 tracks in MK8D, and now we're getting another 4 new characters.

100

There will be over 100 courses in Mario Kart 8 Deluxe by the end of this year.

96 racecourses, and 8 battle courses.

Going back to 32 for 10 would be painful, especially when half will probably be Retro tracks.
 
0
Mario Kart 9 already exists. It's called Mario Kart: Tour, and it's a live service title. MK8DX receiving its tracks shows, and it being connected to NSO, is a model for Nintendo to do live service themselves. Spend 60 bucks for a base number of tracks, pay 50 bucks a year for more tracks on a drip feed.

Not saying MKX can't or won't happen. But I have 0 confidence that we'll get another Mario Kart title anytime in the next half decade.
Or Nintendo suddenly decides that Mario Kart Tour isn’t mainline (like they neglected Mario Galaxy 2 randomly in the 35th Anniversary Direct) and we get Mario Kart 9 as a launch title for Switch 2. Honestly, I dont know if Nintendo is predictable enough to just say „Mario Kart Tour is Mario Kart 9 and we don‘t get a new MK in 5 years“ since its currently the most successful IP and they for sure can just announce a new title.
 
Still disagree honestly. 4K 60 fps TOTK with higher settings like draw distance, would be spectacular even if they didn't do anything fancy like RT.

Especially because it's a game about great heights. Taking in the landscape below and above, in much higher detail and resolution would be a really good showcase.

Happy to agree to disagree on this, and for the record I'm sure ToTK would look great with a simple resolution increase, I just don't think Nintendo was ever planning on using it as the major showcase for the new hardware.

the Series S is a more applicable measure of comparison than PS5

Yes and no. While I do think the existence of the Series S is going to help encourage more third party ports to the Switch 2, those games will still almost always be developed with PS5 as the primary platform and ported down to the Series S rather than vice versa. If publishers are already paying to scale things down to the Series S, then there may be enough overlap to bring down the cost of also producing a Switch 2 version, but fundamentally it's still a PS5 game you're trying to get running on the Switch 2.

It's also worth noting that probably the biggest challenge for porting from PS5 to Switch 2, the CPU, won't be a problem for Series S, as it has basically the same CPU as PS5. Some games may be relatively easy to get running on Series S by reducing rendering resolution and asset quality, but might be near-impossible to port to Switch 2 because of CPU performance differences.

The thing is that the memory controller is baked into the SOC. I expect LPDDR5, same as Orin, but @Look over there has made a reasonable argument that going with 5X might save Nintendo money in the long run - because it's baked into the SOC, going with 5X future-proofs Nintendo against the older standard getting taken over by the newer one in the mobile market.

This is also something they did with LPDDR4X with Mariko, switching from LPDDR4 on TX1 for availability (and probably power draw), even though they weren't getting the speed benefits of LPDDR4X. The Mariko models appeared about a year after the first LPDDR4X phones, where the first LPDDR5X phones have just hit the market. Nvidia also have a LPDDR5X memory controller on Grace, which is likely to leave the factory before Switch 2 does.

That said, my expectation is still for LPDDR5. Even though LPDDR5X might be better in the long run, there would still be higher costs for the first year or two. It's also possible that Nintendo is already planning a mid-gen die shrink anyway, as it may be necessary for a Switch 2 Lite, at which point they could switch to LPDDR5X.
 
Mario Kart 9 already exists. It's called Mario Kart: Tour, and it's a live service title. MK8DX receiving its tracks shows, and it being connected to NSO, is a model for Nintendo to do live service themselves. Spend 60 bucks for a base number of tracks, pay 50 bucks a year for more tracks on a drip feed.

Not saying MKX can't or won't happen. But I have 0 confidence that we'll get another Mario Kart title anytime in the next half decade.

We’ll get a new Mario Kart game within the first 2 years of Switch 2 I’d imagine. It’s an extremely important title in terms of getting people to make the jump. I guess it just depends on how far away you think Switch 2 is!
 
If TotK looks as clean as the 1080p press kit screenshots suggest I'm good until July at least :p .

Also: Metroid Prime Remastered gave me that Metroid Dread/Splatoon 3 feeling again - Switch is awesome if devs know how to stay in spec.
So often developers allow the scope of a game go beyond the capabilities of the hardware they are working with. Prime Remastered showcases just how great a game can look if the art direction is great and the scope is kept in check. Retro is going to have their hands full with making Prime 4 bigger and badder, but also without sacrificing the visuals they achieved with Prime Remastered.

I think its clear Zelda TotK is using the same image reconstruction that Xenoblade 3 used. Perhaps even doing what they did in portable mode for XC3, reconstruct to a higher resolution than output so that they can down sample. Something like render internally at 600p and reconstruct to 1200p with a down sample back down to 1080p for a really clean image.

Ya Prime 1 looks pretty fantastic, even on a 60 inch screen... would look better in 4k but tbh it would be a very diminishing return indeed. Still want it tho lol.

4K was really made necessary because low cost LCD panels kept pushing the market towards larger and larger displays. If you have a 42" TV for example and sit at 7ft away, the benefits are minimal compared to 1080p. Problem is people are purchasing 60" TV's and still still sitting that close, and that's when the need for the higher resolution comes into play.

Prime Remastered looks very clean. I was surprised that it renders at 900p and not 1080p, it looks that clean. Apparently they have implemented 2X MSAA. I wish more of Nintendos teams saw the value in quality AA.
 
I just don't know what the next Mario Kart could possibly do to separate itself from MK8D.
New tracks with new kinds of gimmicks, new powerups, new kinds of vehicles to go along with kart/bike/ATV, additional modes 8 didn't have like DS's mission mode, better integrated online modes, the sky's the limit.
 
I just don't know what the next Mario Kart could possibly do to separate itself from MK8D. I mean we're going to have 70 tracks in MK8D, and now we're getting another 4 new characters.
It’s ok. We’ll leave it up to the developers at Nintendo to show you what they’ll do to separate it from MK8D.
 
It’s ok. We’ll leave it up to the developers at Nintendo to show you what they’ll do to separate it from MK8D.
It's a matter of content really. Even with new gimmicks or mechanics, the content itself will pale in comparison to MK8D. Let's say they go the Diddy Kong Racing route in terms of new vehicle types during a race. Would that be worth missing out on over 60+ tracks?

Smash is suffering from the same problem. A "reboot" of the franchise with new mechanics and movesets for the core ~24 characters could be nice at first - then the extreme lack of fighters and stages relative to the previous title starts to become very apparent.

I'm sure mechanically the new MK (and Smash eventually) will be fun and novel. But there's a good chance both games will be lacking a significant amount of content from their previous titles.
 
Apparently the last TotK's trailer is compatible with 5.1 surround sound using Youtube. And I wonder... Is Nintendo Switch compatible with this kind of audio?
This is confirmed by the japanese Twitter account from Zelda.
@ZeldaOfficialJP
 
Apparently the last TotK's trailer is compatible with 5.1 surround sound using Youtube. And I wonder... Is Nintendo Switch compatible with this kind of audio?
This is confirmed by the japanese Twitter account from Zelda.
@ZeldaOfficialJP
Prime Remastered has it.
 
It's a matter of content really. Even with new gimmicks or mechanics, the content itself will pale in comparison to MK8D. Let's say they go the Diddy Kong Racing route in terms of new vehicle types during a race. Would that be worth missing out on over 60+ tracks?
I'm going to get the next 3D Mario platformer, even if it includes 0 stages from 3D World or Odyssey. What's new is a more important hook, and what's old they'll probably drag out through a new DLC/subscription like they are doing now.
 
Some folks in the thread I posted yesterday believe there will be a bigger gap between Switch 2/PS5 than there was between Switch/PS4.

Is this based on anything concrete? I don't get it.
There dont will be bigger gap, Switch 2 will be have really strong SoC, power will be same or similar to Series S, and thanks to Series S, there its possibility that almost all new AAA games will be ported to Switch 2
 
There dont will be bigger gap, Switch 2 will be have really strong SoC, power will be same or similar to Series S, and thanks to Series S, there its possibility that almost all new AAA games will be ported to Switch 2

I think you shouldn't expect series S power level.
It would be nice if we get ps4 pro power level (with DLSS on top of that).
 
Please read this staff post before posting.

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


Back
Top Bottom