• 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!)

Frame generation really isn't a benefit below a certain base framerate, since your input latency is still tied to the native frames.

I guess then it's really just probably a very low native res using DLSS and efficiency gains in UE5 doing a lot of the heavy lifting.
 
0
Well, when you think about it, for a tech demo input latency isn't exactly a major consideration.
Yes, but it would be an extreme misrepresentation to tell the devs you're courting "This is running on target specs!" and then later reveal it was running at 15fps with frame generation. Not that I believe that was the case, but I hope you can see what I'm saying.
 
Should the new Switch be able to 4k/60fps (DLSS) all Switch games that are 720p/60fps if patched and the only difference between versions is resolution?
If patches are possible I wouldn't expect anywhere near most games to be affected, but rephrasing your question to "Should any existing 60fps Switch game be able to be a 4K60 Switch 2 game if patched/ported to take advantage of the new hardware?" I can't think of a reason why not. It would be more easily done for some games than others.
Yep! I personally imagine MP4 would target similar performance and resolution as MPR. Up to 900fps and targeting 60fps on Switch docked (600-720p on handheld). Didn't XB3 and Totk use some kind of upscaling FSR too btw?
Xenoblade 3, at least, used a temporal upscaler. So closer to FSR2, though not precisely that.
Ghostsonplanets quoting Digital Foundry said:
The Legend of Zelda: Breath of the Wild's mooted tech demo is an altogether different proposition, sounding more like a practical example of how DLSS can be transformational for consoles - but again, context is king and we don't have any. Our understanding is that the demo runs at 60 frames per second with a 4K output target, but with DLSS's various quality levels, the native internal resolution could be anything from 720p upwards (Switch 1 runs at 720p30 in mobile mode, 900p30 docked).
There's nothing magical about the modes they gave names for easier choosing by consumers, so the demo could theoretically have an internal resolution of like 10p. Except I imagine some of the reports would then mention how shit the new version of BOTW looked.
Frame generation really isn't a benefit below a certain base framerate, since your input latency is still tied to the native frames.
Well, when you think about it, for a tech demo input latency isn't exactly a major consideration.
The fact that frames further apart in time makes the created frame look worse still is a consideration, even if Nintendo somehow got frame generation working only for bullshot demo presentation purposes.
 
So, a few general comments on the updates:

Firstly, and most importantly, now that "around March" is on the table, I'd like to officially announce the re-launch of #TeamLeapDay. We'll be accepting membership applications by messenger pigeon, which must be accompanied by a one thousand word essay on why Wave Race should be a launch title.

Secondly, it looks like I wasn't completely wrong about BoTW, with both DLSS and fast loading confirmed to be in the demo. No mention of ray tracing, but two out of three ain't bad.

Actually being able to run DLSS at 4K/60fps is a bit better than I'd expected. My personal expectations were 4K/30fps, or around 1440p/60fps, just due to the cost of running DLSS itself, but it looks like it's either cheaper than I expected, or the hardware is more capable than I expected, or a bit of both. Good news either way.

The talk of "instant" load times in BoTW is interesting. Of course nothing is truly instant (even Ratchet & Clank uses a short portal sequence to hide loading, a bit like the old door-opening trick on Metroid Prime back in the day), but we're probably talking around a second or so for it to feel instant. That's about a 30x speed up compared to the Switch.

It's important to note that there are multiple factors which impact loading speeds, not just the speed of the storage itself (although as I've explained before, I was already expecting that to be much quicker). Roughly speaking, I'd categorise them into three different bottlenecks which could factor into how fast loading is:
  1. The speed of the storage medium itself. Obviously this can be a big bottleneck, as if you want to pull 1GB of data off a storage device which maxes out at 100MB/s, you can't do that any quicker than 10 seconds.
  2. The CPU overhead associated with pulling that data off the storage medium. The main part of this nowadays is decompression. If your CPU can't decompress data fast enough, then even if you've got super fast storage you're not going to be leveraging it (unless you want to ship purely uncompressed assets and balloon the game size). Aside from decompression, though, there's still some additional overhead of just communicating with the storage medium and transferring the data, which can be non-trivial.
  3. Non-storage related CPU work. This is doing everything you need to do to actually set up the game state. Allocate memory for all the different things you need to keep track of during gameplay, initialise all the different systems which are going to be running, and sync the game state with the save file. This is also where you're doing any kind of procedural generation, which is often overlooked when thinking about load times. People playing No Man's Sky on the PS5 for the first time were really puzzled why it wasn't loading much quicker than the PS4, but the game is fully procedural, so it was never really bottlenecked by storage in the first place.
On Switch, it seems most games are bottlenecked by number 2. We know that T239 contains a dedicated File Decompression Engine, so ideally we're not bottlenecked here anymore, but if we were, it would suggest that the FDE is around 30x as fast at decompressing data as the Switch CPU. I don't know if we can say for sure what data rates BoTW was operating at while loading, but I don't believe it's noticeably faster on internal storage or microSD than game cards. Game cards top out at 50MB/s, so if the Switch CPU was able to keep up with that, then we'd be looking at 30x as much, or about 1.5GB/s of compressed data coming into the FDE. It could certainly be less than that, with the CPU perhaps bottlenecking things even lower on the Switch, but honestly 1.5GB/s or so wouldn't be crazy. Both MS and Sony designed much faster decompression hardware 3 years previously, and Nvidia isn't exactly a slouch when it comes to designing fast coprocessors.

If we were bottlenecked on storage, then again comparing to the original Switch game card as the baseline we'd once more be looking at around 1.5GB/s. For a while I've been considering UFS 3.1 to be reasonably likely, which ranges from 1.7GB/s to 2GB/s read speeds, and is pretty much in line with this. In fact, if Nintendo actually is considering 512GB of storage for the console, then 1.7GB/s UFS 3.1 is the slowest option they would have. Nobody makes 512GB modules for either eMMC or UFS 2.

The last one probably isn't trivial for BoTW on the original Switch, although likely a lighter workload than it is for ToTK with its increased number of gameplay systems. I definitely don't expect a 30x speedup here, but between taking the decompression work off the CPU, and just having more, faster CPU cores to do the work, it should be much faster either way.

Of course this is all based on a very specific reading of "instant" loading. If 2 seconds counts as instant, then divide every number in half. If it's 0.5 seconds, multiply by 2, etc. However, given expectations of internal storage close to 2GB/s, and dedicated decompression hardware to remove the main CPU bottleneck, dropping load times from 30s to around 1s seems quite reasonable.
It depends on your definition of "noticeably," but the Switch's internal storage is a little faster than SD cards which are somewhat faster than game cards, with up to several seconds separating them overall in loading sequences. It's certainly noticeable to speedrunners.
 
Last edited:
Yes, but it would be an extreme misrepresentation to tell the devs you're courting "This is running on target specs!" and then later reveal it was running at 15fps with frame generation. Not that I believe that was the case, but I hope you can see what I'm saying.
Oh I absolutely do, I was being a bit cheeky. That would be a way to "cheat" in a tech demo.
 
Still thinking it will be something like this:

March - Reveal
June - Games Blowout
September or after (2024) - Release

If Nintendo do decide to go earlier though then I could see it being in May.
 
If patches are possible I wouldn't expect anywhere near most games to be affected, but rephrasing your question to "Should any existing 60fps Switch game be able to be a 4K60 Switch 2 game if patched/ported to take advantage of the new hardware?" I can't think of a reason why not. It would be more easily done for some games than others.
Xenoblade 3, at least, used a temporal upscaler. So closer to FSR2, though not precisely that.
It was mentioned here a week ago that data miners apparently found evidence of enhancement support added to one of the Xenoblade titles following a patch update, then of course there was PokéPoster asserting work being done on Scarlet & Violet, and now an improved Breath of the Wild showcased at Gamescom

Probably safe to say if it's a (recent) Pokémon, Mario, Zelda, Splatoon, Kirby or Xenoblade title, it's getting a makeover
 
Well, when you think about it, for a tech demo input latency isn't exactly a major consideration.
Artefacts would be, though, and I can imagine them being rough trying to use DLSS-FG to get from 15fps to 30fps. That's a big old gap between each frame you're trying to interpolate over.

Anyway, the issue with frame generation is that it's not free performance. Even if you ignore the Ada Lovelace requirement and the OFA, etc., the actual work of DLSS-FG is done on tensor cores, and if that tensor core work takes longer than it does to actually render the frame, then frame generation does nothing for you. It's also run at the output resolution of DLSS's scaling, so if Nintendo were to (say) run a game at 720p internal and scale it up to 4K via DLSS ultra-performance mode, it would have to do the same amount of work on interpolating those 4K frames as an RTX 4090 does on interpolating native 4K frames.

We looked at this a while back, and although there aren't any hard numbers on it, Oldpuck estimated around 45ms for T239 to run frame generation at 1440p, so you can probably double that at 4K. Unless Nvidia can massively improve performance somehow, it just doesn't seem to make sense for a low-power device like the Switch.
 
The commit is from November 2022? Or was that not found before
The commit is from August 2022, when that file was initially created. And we did discuss it in the thread, probably in September 2022.

Anybody know what the difference would be in "ampere_A" and "ampere_b" and if it matters for us?
A is the first sub-family (there's probably an official Nvidia term but it's escaping me at the moment) and B is the second. For Ampere, the first was datacenter and the second was everything else.
 
Anybody know what the difference would be in "ampere_A" and "ampere_b" and if it matters for us?

Ampere A is the A100 HPC GPU, which is quite different from the gaming cards (eg support for FP64, etc.).

Honestly, this doesn't mean a whole lot. We've known that T239 is Ampere for over a year, and Orin (T234) already deviates from standard desktop Ampere chips like the RTX 3080 in meaningful ways (eg double size tensor cores) yet is still listed as Ampere_B here.
 
All you nerds talking about your ray traced metroids and Mario's

Meanwhile I'm dreaming of games that look like this!



I wonder how hard/easy it will be to get DLSS/RTX into NSO games, because that would be a cool way to resfresh the older games. I would like to see how it would work with Gamecube/Wii games. Also, the tensor cores should be able to help create ai generated improved textures right?
 
Anybody know what the difference would be in "ampere_A" and "ampere_b" and if it matters for us?
Ampere A is the HPC chip. Has no rt cores and more tensor cores. Ampere B is the gaming design

How far we’ve come. Remember the dark days when Capcom said Wii couldn’t run the title screen for Resident Evil 5?
And then when the TX1 was announced for Switch, people posted the poor performing shield port proving Switch was doomed to fail
 
I wonder how hard/easy it will be to get DLSS/RTX into NSO games, because that would be a cool way to resfresh the older games. I would like to see how it would work with Gamecube/Wii games. Also, the tensor cores should be able to help create ai generated improved textures right?
Probably have to take the source code and attach a new renderer to it. That's how Mario 64 RT is doing it
 
Someone gotta call Tokyo Big Sight and ask them if the venue is available the Friday before for an event, I can actually ask my Japanese girlfriend to do it. Shall we ?
That’s a generous offer. I have a family member who’s a professional conference organizer. From the bits and pieces I heard, it seems unlikely that a conference center would provide you with their booking information unless you’re an organizer. Also, setting up a event of that size, it probably will require a few days to set up. So even if it’s confirmed that Nintendo’s booked the South Halls on the days prior, we can’t be certain of a press conference.

Edit: typo
 
Last edited:
Of course this is all based on a very specific reading of "instant" loading. If 2 seconds counts as instant, then divide every number in half. If it's 0.5 seconds, multiply by 2, etc. However, given expectations of internal storage close to 2GB/s, and dedicated decompression hardware to remove the main CPU bottleneck, dropping load times from 30s to around 1s seems quite reasonable.
This feels like something one of the birds could clarify. The fade in and out animation even if it doesn't drop to the BotW loading screen could give you a few seconds to load out and in objects.

This line of thought experiment makes me wonder how large a world could the Portal engine support with some of the modern disk enhancements.
 
0
I tend to be on the skeptical side of things when it comes to Nintendo hardware, but that tech demo sounds seriously impressive. DLSS sounds like its doing some real heavy lifting, so I suppose that Nintendo-NVidia partnership for the past decade or so has really paid off.

The upscaled 4k sounds great, but truthfully its the frame rate that really sounds like gravy to me. I can't imagine I'd really need much else as long as everything can hit that 1080p / 60fps threshold, DLSS and 4k just makes it all the sweeter.

I've always been team Q2/Q3 2024 (sometime around June or July), so my personal guess is that the March 2024 date floating around is probably a reveal date instead of a launch date. My guess is that the reason they're throwing it out there is to see if they can get any of these smaller developers (because I'm confident your big companies like Bandai Namco, Capcom, Square Enix, etc. have already known about the Switch 2 for a minute now) to see if they can do anything within the next couple of months for launch title / first year releases.
 
It was mentioned here a week ago that data miners apparently found evidence of enhancement support added to one of the Xenoblade titles following a patch update, then of course there was PokéPoster asserting work being done on Scarlet & Violet, and now an improved Breath of the Wild showcased at Gamescom

Probably safe to say if it's a (recent) Pokémon, Mario, Zelda, Splatoon, Kirby or Xenoblade title, it's getting a makeover
Could you find the original post/source for this? I'm interested in seeing this.
 
I tend to be on the skeptical side of things when it comes to Nintendo hardware, but that tech demo sounds seriously impressive. DLSS sounds like its doing some real heavy lifting, so I suppose that Nintendo-NVidia partnership for the past decade or so has really paid off.

The upscaled 4k sounds great, but truthfully its the frame rate that really sounds like gravy to me. I can't imagine I'd really need much else as long as everything can hit that 1080p / 60fps threshold, DLSS and 4k just makes it all the sweeter.

I've always been team Q2/Q3 2024 (sometime around June or July), so my personal guess is that the March 2024 date floating around is probably a reveal date instead of a launch date. My guess is that the reason they're throwing it out there is to see if they can get any of these smaller developers (because I'm confident your big companies like Bandai Namco, Capcom, Square Enix, etc. have already known about the Switch 2 for a minute now) to see if they can do anything within the next couple of months for launch title / first year releases.
A reveal date for March 2024 wouldn't make a lot of sense, if it's entering production in December then descriptive and photographic factory leaks will follow within days of beginning assembly - if Nintendo don't announce it before mass production then they'll lose control of the information cycle
 
0
There's a chance they have a special version of DLSS running but I can't imagine they're doing something like running the demo at 240p internally and using DLSS to boost it to 4k. That would either be a mushy muddy mess or if it did work well, actual wizardry that NVidia would want out on their PCs instead of locked to a business partner's console. UE5 has become a lot more stable and a lot more optimized over 2 years. That's the most likely explanation.

Just like there are efficiencies and improved performance on Switch by using the low level API NVN, I would think the same will hold true for DLSS and Ray Tracing for NVN2 on SNG. When Nvidia implements DLSS onto SNG, they will have low level access to the hardware, and know exactly what that hardware is. Just looking at a PC graphics card with similar specs wont be like for like because on PC the API will not be low level. If you look at shader performance on Switch using OPEN GL, its going to be notably worse than when using NVN. A good showcase of this is Sniper Elite 3 and 4. Rebellion games has interviews out there where they express just how important it is to use NVN to get good performance from Switch and there titles really do shine well on Switch. If shader performance improves substantially with low level access, I see no reason that DLSS and Ray Tracing wont see similar benefits.
 
Could you find the original post/source for this? I'm interested in seeing this.
Yes, I've partly misremembered it, it was to do with the OS but the comment was made in response to one discussing Xenoblade:

Datamines of the OS show the existence of a so-called "datapatch", which is speculated to be for next gen patches. It was added to the OS around the same time as the new memory management system, which is also for new hardware, so I think it probably is to do with next gen patches.
 
Yes, I've partly misremembered it, it was to do with the OS but the comment was made in response to one discussing Xenoblade:

This is "nn::ncm::[sic]DataPatchId".
This was added with [15.0.0+].


The Switch 15.0.0 system update was released on October 11, 2022 (UTC). This Switch update was released for the following regions: ALL, and CHN.

No game has used "datapatch", ordinary "patch" seems to be what games, even with big updates like Splatoon 3, use.

I believe it may have something to do with redirecting file requests, which would make sense for a next gen patch with new assets.
 
Every new piece of info makes me want to see this thing more and more. At this point, just give me a price and a release date so I can make sure I've got the money Day 1!
 
I think Game Freak is sweating hard right now, they will have to scale up from their gamecube style graphics hard on the next switch, maybe go on a hiring spree for new devs?
Game Freak hire more devs? Hell will freeze over before this is a reality. 3 people and a cat will continue to make Pokemon until the end of time.
 
I think Game Freak is sweating hard right now, they will have to scale up from their gamecube style graphics hard on the next switch, maybe go on a hiring spree for new devs?

They have been putting their devs on a Unreal Engine intensive class for a while now
 
I think Game Freak is sweating hard right now, they will have to scale up from their gamecube style graphics hard on the next switch, maybe go on a hiring spree for new devs?
They're currently on one, and with Scarlet and Violet DLC (and a likely next gen patch) spread out over 2023 and 2024, it looks like they might be taking a bit of extra time.
 
0
I think Game Freak is sweating hard right now, they will have to scale up from their gamecube style graphics hard on the next switch, maybe go on a hiring spree for new devs?
The Game Freak situation isn't salvageable without Nintendo stepping in to start making demands or inserting themselves into the heart of the development process, which they probably can't do due to whatever way the contract was drawn up or else they would've done so a long time ago already

It's unfortunate, but we're at a point now where mainline Zelda is outselling mainline Pokémon, so whatever reputational damage is being done by Game Freak is thankfully being more-than-undone by the pure prestige that is Production Team 3's ability to bring out the absolute best of the hybrids, which they will no doubt continue doing into the NG era
 
0
Assuming BC is a given, which I think it is, what is the current consensus on how unpatched next Gen games will work on it? I'm mostly referring to those with dynamic res or fps targets or sloppy fps such as age of calamity, trials of mana or Xenoblade 2.

Will those work better simply by virtue of being on a superior hardware?
 
There's a chance they have a special version of DLSS running but I can't imagine they're doing something like running the demo at 240p internally and using DLSS to boost it to 4k. That would either be a mushy muddy mess or if it did work well, actual wizardry that NVidia would want out on their PCs instead of locked to a business partner's console. UE5 has become a lot more stable and a lot more optimized over 2 years. That's the most likely explanation.
I think I've read in this thread speculation that Ultra Performance has been built with Switch 2 in mind.
 
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