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

[Eclissi]

I never understood why it is so complicated for fake insiders to understand that inserting the name and price of the still undisclosed console into an invented leak makes the fake clear. These are data that have not yet been defined a year before marketing.

 

[ertaboy356b]

That registration for gold points can be a very restricted option. For context: "Gold Points must be claimed within one year of the game’s original release date in the Nintendo eShop. If a physical version of a game is released later than the digital version, it may no longer qualify for Gold Points."
Not to mention all the people who took years to find this out lol (many still don't know about it)

And even then, I don't think Nintendo would like to see the tsunami of people reselling their physical copies because they now have its digital licenses.

For that, I can see Nintendo opting for a physical-to-digital trade-up program at a worldwide level.

But, anyway, I don't believe they would go with digital only on Switch 2.

Yeah, here in Asia in order to claim gold coins, you need to be in the same region the game is. The problem is, games sold here are on mixed regions. Some are NA, some are JP, some are ASIAN (HK) so yeah. For example, the physical version of Etrian Odyssey needed you to set your region to HK before you can claim gold coins or the free DLC.

 

[SoldierDelta]

It's doable on Switch. We've gotten a ton of Wii U ports already. The biggest challenge will be removing the Wii u pad stuff.

I think the biggest factor of a switch port coming is more so the timing. There's been at least 2 years between every new Xenoblade game on switch, and that counts when you include DlC stories like Torna and FR

-XB2 &XB2 Torna 2017 & 2018
-XB1 DE 2020
-XB3 & XB3 FR 2022 & 2023

So it's safe to predict XBX would be coming out in 2025. Maybe end of 2024 if we're lucky. Tbe demand by fans for X is there. It's even more likely on switch if Switch 2 doesn't come out until Q4 2024 and switch sales are going relatively strong

I was mostly talking from what I've heard. Xenoblade X is a fucking huge game that only just ran on the Wii U, and even then it needed patches and dlc to be able to run properly for the most part. For the sake of argument though, let's presume that Monolith Soft does was it to be a cross-gen game that has enhancements on the Switch 2 but can feasibly run on the Switch 1. The dates you listed sound about right.

It's worth pointing out that, thanks to Future Redeemed's radio, we know that Monolith Soft are happy to work on their previous titles in the Xeno metaseries (including the elusive Xenoblade X port). Xenogears is kind of a crapshoot that relies on Square Enix waking up on a certain side of the bed (maybe if Monolith Soft is given the IP to work on we can expect a remake at some point, idk), but we can assume that Xenosaga is very likely. Hell, the trademark for those games were in the information for Future Redeemed and the ports of the Baiten Kaitos games are a month away. If they're ports, I predict mid-2024. If they're remasters, we can expect them 2025... maybe 2026. Regardless we're getting them long before a potential "Xenoblade 4" (I don't expect a XC4 until 2027 at the very very very earliest).

 

[ILikeFeet]

I think we're overselling the Xenosaga stuff. Outside of pulling ideas for the next Xenoblade, the chances of them working on a non-nintendo IP, let alone 3 games of them, is very low

 

[grobla123]

re: reddit bullplop, "Thumbsticks have their own visible buttons", what would that mean, just like a coloured lettered button for clicking L3/R3 or like multiple buttons on each stick (what would that work like?).

Kid Icarus game reference right at the end after fairly innocuous stuff. Why they gotta do me like that.

 

[SoldierDelta]

Designing a cart slot that accommodates multiple cart formats really isn't that big a deal and is something Nintendo has done multiple times in the past.

This is honestly a very fun discussion simply due to how many times it has happened.
If those "3d-nand" rumours are to be believed (Idk how confirmed that is), I can imagine the carts to be made in a nigh-identical form factor to the Switch 1 carts. Literally all you'd need to do is have a software check saying "Hey, is this system able to run these things?" and then you're good to go.

 

[TLZ]

Man I can't wait til the console leaks and emotions are all over the place.

"Inject it straight into my veiiiinzzz"
"Lel wth is that name"
"Day 1"
"DOA"
"Lol nintendoomed"
"Wut"
"But can it run crysis?"
"Pro when?"
Etc etc

Want that day now!

 

[SoldierDelta]

I think we're overselling the Xenosaga stuff. Outside of pulling ideas for the next Xenoblade, the chances of them working on a non-nintendo IP, let alone 3 games of them, is very low

I sorta disagree. For one, Monolith Soft doesn't need to work on those games for them to be ported, Bamco seems happy enough to port them in their own time. The second point that could be made is that Bamco seems very happy to work with Nintendo, as we've seen time and time again. It could literally just be a co-production between Monolith Soft and Bamco and Nintendo's involvement can be limited to "Hey, let Takahashi cook" and nothing more.

That being said, other Xeno stuff is very much a crapshoot. The odds of Xenogears leaving the Square Enix archives is very low, but a man can dream.

 

[Evil Lucario]

I was mostly talking from what I've heard. Xenoblade X is a fucking huge game that only just ran on the Wii U, and even then it needed patches and dlc to be able to run properly for the most part. For the sake of argument though, let's presume that Monolith Soft does was it to be a cross-gen game that has enhancements on the Switch 2 but can feasibly run on the Switch 1. The dates you listed sound about right.

It's worth pointing out that, thanks to Future Redeemed's radio, we know that Monolith Soft are happy to work on their previous titles in the Xeno metaseries (including the elusive Xenoblade X port). Xenogears is kind of a crapshoot that relies on Square Enix waking up on a certain side of the bed (maybe if Monolith Soft is given the IP to work on we can expect a remake at some point, idk), but we can assume that Xenosaga is very likely. Hell, the trademark for those games were in the information for Future Redeemed and the ports of the Baiten Kaitos games are a month away. If they're ports, I predict mid-2024. If they're remasters, we can expect them 2025... maybe 2026. Regardless we're getting them long before a potential "Xenoblade 4" (I don't expect a XC4 until 2027 at the very very very earliest).

I'm pretty sure the data packs were only recommended because data transfers off discs is just awful, so they needed installs instead. I think a Switch port could bypass that, even on an SD card.

However, once all's said and done, Xenoblade X actually runs pretty well on Wii U with very few framerate dips for the most part. But X cheats a lot with making it run, like how every NPC has no collision detection, vegetation isn't as robust, not as much modern techniques you see in 2, DE, and 3, and more. It also has way worse pop-in than 2.

Now that being said, it might come down to the fact that Xenoblade X's engine was so built for the Wii U, porting it to Switch architecture would be expensive. Kinda like a PS3 situation but not as bad. But also, Xenoblade X has the biggest file size out of any Nintendo game, above 20 or so GB. So shipping it on a cartridge would likely be even rougher.

 

[MuchLonk]

Watch them sell a Switch card reader, like the GBA card reader on GameCube. /j

 

[SoldierDelta]

I'm pretty sure the data packs were only recommended because data transfers off discs is just awful, so they needed installs instead. I think a Switch port could bypass that, even on an SD card.

However, once all's said and done, Xenoblade X actually runs pretty well on Wii U with very few framerate dips for the most part. But X cheats a lot with making it run, like how every NPC has no collision detection, vegetation isn't as robust, not as much modern techniques you see in 2, DE, and 3, and more. It also has way worse pop-in than 2.

Now that being said, it might come down to the fact that Xenoblade X's engine was so built for the Wii U, porting it to Switch architecture would be expensive. Kinda like a PS3 situation but not as bad. But also, Xenoblade X has the biggest file size out of any Nintendo game, above 20 or so GB. So shipping it on a cartridge would likely be even rougher.

Okay, this is a very nice explanation, thank you.

If I had to make a solid guess as to why it hasn't been ported then... it's just because Monolith Soft has pretty much been working non-stop since the Switch launched. The fact that Xenoblade 2, Torna, DE and FC, 3 and Future Redeemed all launched in 5 years with Monolith Soft doing support work on Splatoon, Animal Crossing and both Open-World Zeldas is absolutely nuts.

If Xenoblade X is to be ported at all (which is very likely imo), it will be around 2025. Monolith Soft probably won't release anything next year with the exception of support-work stuff (even then I doubt it).

 

[Samuspet]

I'm curious where you got a 5x CPU power gap between PS5 and Switch 2/Drake? Switch 2 could actually close the power gap a bit from switch vs PS4 (which was about 3.5x in speed).

1.5tflop mobile version of Switch 2 (which as we know from Switch games the games are built around the limitations of mobile mode then scaled up for docked mode) versus 9-10tflop PS5 GPU. PS5 ram bandwidth and SSD speeds are absolutely insane too. I'd say overall PS5 will be around 5x more powerful than Switch 2 but that's a hell of a lot better than the Wii which was 40x weaker than PS3 then Wii U which was 20x weaker than PS4 then Switch which was 10x weaker than PS5.

 

[MusicProposition]

Did either one of those platforms have Nintendo's undivided attention? You also admitted the Pokemon games weren't even the mainline games. Those platforms weren't successful due to droughts they had due to Nintendo's split resources. Give N64 all of support provided to Gameboy Color and all of a sudden N64's library looks far superior along with Nintendo fans no longer having a choice in how to consume Nintendo's ip. Gamecube gains Gameboy Advances support. You realize I'm not saying just give N64 and Gamecube all of the games provided to Gameboy Color and Gameboy Advance. I'm saying all of the support those handhelds received would have instead been spent on building games for those home consoles. That's the difference between them and Switch.

I see your point, but you make it seem like this thing - Nintendo’s undivided attention - is their key to success, and I just think there’s more to that. The Switch worked out very well, but you can’t do that over and over again and they have to get many things right to be successful. Not that they couldn’t, but I don’t think it’s gonna be easy either. Also, while they consolidated their resources, you can’t argue that development took them much longer over time. We’re not gonna get games like Majoras Mask made in just a year.

 

[Evil Lucario]

Okay, this is a very nice explanation, thank you.

If I had to make a solid guess as to why it hasn't been ported then... it's just because Monolith Soft has pretty much been working non-stop since the Switch launched. The fact that Xenoblade 2, Torna, DE and FC, 3 and Future Redeemed all launched in 5 years with Monolith Soft doing support work on Splatoon, Animal Crossing and both Open-World Zeldas is absolutely nuts.

If Xenoblade X is to be ported at all (which is very likely imo), it will be around 2025. Monolith Soft probably won't release anything next year with the exception of support-work stuff (even then I doubt it).

Yeah, with the progression of Xenoblade on Switch, it feels like Monolith wanted to focus on their mainline stuff first out of branding and momentum, and X simply didn't fit into those plans just quite yet. Which makes sense even development-wise, like you can see how making DE was basically a stealth partial asset farm to help out 3 and FR's development.

 

[SoldierDelta]

Yeah, with the progression of Xenoblade on Switch, it feels like Monolith wanted to focus on their mainline stuff first out of branding and momentum, and X simply didn't fit into those plans just quite yet. Which makes sense even development-wise, like you can see how making DE was basically a stealth partial asset farm to help out 3 and FR's development.

I still stand on the hill that Takahashi is playing the longest game ever. Dude is allowed to work on his passion project that failed twice due to corporate shenanigans and is finally able to make it in exchange for making Nintendo billions via working on Zelda, Animal Crossing and Splatoon. Real King behaviour.

Regardless, Xenoblade X is seemingly not canon for now so it's not ""needed"" in the same way something like Xenoblade 1 was needed in the lead up for Xenoblade 3/FR. It'll be relevant eventually, just not now.

 

[Instro]

Xenoblade X always seemed like a better bet for the Switch successor, considering how it pushes pretty much every aspect of the WiiU to the edge. That said it's also in an odd position where we've heard several times that development of the game was notably difficult on the team. They may not be eager to relive that particular chapter of the studio's life.

 

[ILikeFeet]

Xenoblade X always seemed like a better bet for the Switch successor, considering how it pushes pretty much every aspect of the WiiU to the edge. That said it's also in an odd position where we've heard several times that development of the game was notably difficult on the team. They may not be eager to relive that particular chapter of the studio's life.

Then let some other team work on it. They already adapted the engine to work on new hardware and upgraded the renderer. So let an external studio do the asset work while a different staff at Monolith integrate it.

 

[ShadowFox08]

1.5tflop mobile version of Switch 2 (which as we know from Switch games the games are built around the limitations of mobile mode then scaled up for docked mode) versus 9-10tflop PS5 GPU. PS5 ram bandwidth and SSD speeds are absolutely insane too. I'd say overall PS5 will be around 5x more powerful than Switch 2 but that's a hell of a lot better than the Wii which was 40x weaker than PS3 then Wii U which was 20x weaker than PS4 then Switch which was 10x weaker than PS5.

In your original post that I quoted, you said 3x in GPU and 5x in CPU in regards to the power gap with PS5 and Switch 2.

If it's like the Switch, CPU clockspeeds are unlikely to change between handheld and docked mode. If we get 8 A78 CPU cores, with 7 for gaming at a relatively modest 1.5Ghz per CPU core, we're looking at possibly 2.3x difference in speed maybe (PS5 being 3.5-3.6Ghz). But I'm spit balling, since AMD Zen 2 CPU cores performance per Hz is more performant than ARM A78 in single core performance metrics.

We'll see if there's a 5x gap in performance overall. I think the overall gap will be smaller than switch vs PS4. I'm not worried about SSD speed. RAM Bandwidth and CPU will be the main gaps. But if we have more RAM available than X series S, we're probably gonna be fine in that department. CPU will be interesting.

 

[lemonfresh]

Man I can't wait til the console leaks and emotions are all over the place.

"Inject it straight into my veiiiinzzz"
"Lel wth is that name"
"Day 1"
"DOA"
"Lol nintendoomed"
"Wut"
"But can it run crysis?"
"Pro when?"
Etc etc

Want that day now!

If the skip out on the oled screen it is most certainly not a "day 1" for me

 

[HankyPanky]

they aren't even trying with these 'leaks'. of course 4k is going to be a feature that's what DLSS is there for.

 

[Gerald]

In your original post that I quoted, you said 3x in GPU and 5x in CPU in regards to the power gap with PS5 and Switch 2.

If it's like the Switch, CPU clockspeeds are unlikely to change between handheld and docked mode. If we get 8 A78 CPU cores, with 7 for gaming at a relatively modest 1.5Ghz per CPU core, we're looking at possibly 2.3x difference in speed maybe (PS5 being 3.5-3.6Ghz). But I'm spit balling, since AMD Zen 2 CPU cores performance per Hz is more performant than ARM A78 in single core performance metrics.

We'll see if there's a 5x gap in performance overall. I think the overall gap will be smaller than switch vs PS4. I'm not worried about SSD speed. RAM Bandwidth and CPU will be the main gaps. But if we have more RAM available than X series S, we're probably gonna be fine in that department. CPU will be interesting.

im still hope for 2GHz for CPU

 

[thesillyfellaelde]

Yeah, with the progression of Xenoblade on Switch, it feels like Monolith wanted to focus on their mainline stuff first out of branding and momentum, and X simply didn't fit into those plans just quite yet. Which makes sense even development-wise, like you can see how making DE was basically a stealth partial asset farm to help out 3 and FR's development.

In my eyes it's also because they felt like it wasn't the right time power wise. Monolith Soft's moto ever since X is to build up a good large library of assets they can use for their games, animations, textures, models, rendering tech, particles, effects, and so on. Xenoblade X is codenamed Spacetravel, and we know that early on they wanted to have spaceships you could travel in (also seen in the Xenoblade X concept art book) and multiple planets you could visit. Just like how a fully realized open world couldn't be done on the Wii for Xenoblade 1, they instead saved the idea for the more beefier console in the future. I think this is probably the same case for X2 and if they're aiming to make it the most ambitious Xeno game to date, where one of the big reasons might be wanting to wait for the next jump in power so that they can go all out in crafting a unique and even loftier open world game for its sequel. Instead of working on making that for a console that's essentially a portable Wii U (at the very least a console that wasn't as big of a jump in power as Wii to Wii U) and then having to make sacrifices in the game's scope to fully get it working, it worked in their favor to embrace the numbered games fully first with their more limited worlds, build up the IP and brand recognition, keep building on a large library of assets that they can streamline game development with and focus on making more new things for games instead of remaking old things, finishing up one story arc first instead of jumping between 2, and finishing off the trilogy just in time to make sure that the next Xenoblade game will be targeted to fully embrace new hardware and be focused on using that new leap in power to evolve the franchise further.

Additionally Xenoblade X also has some online play in its concept, so its interesting to think how much will be put into the online play for X2 if it's going to be expanded on. The online play of X I felt was a stepping stone for them, so I'm very interested to see how its going to be evolved and make it even more appealing to pick up X2.

Next Xenoblade X is gonna be X2 I definitely think, dont forget that Spacetravel is the game's codename, and one big part of Future Redeemed's radio is talking about spacetravel Future Redeemed is one big glimpse into the future of the series as stated by Takahashi, and what the game features is a whole bunch of things from X conceptually, not to mention the original Ares That's another point actually, I fully believe X is connected to the entire story and has a big juicy backstory for everything, and what happens in the Xenoblade trilogy is integral to understanding what's going to happen in X.

 

[TLZ]

If the skip out on the oled screen it is most certainly not a "day 1" for me

Then you fall under

"OLED when?"

 

[Mr Swine]

If Switch 2 has a 1080p screen how is Nintendo going to avoid artifacting and other weird scaling issues when games run 720p in undocked mode?

 

[Adulfzen]

If Switch 2 has a 1080p screen how is Nintendo going to avoid artifacting and other weird scaling issues when games run 720p in undocked mode?

from Nintendo POV there is probably not enough consumers that will use BC and also care about said artifacting to make it a problem worth solving (not saying they won't try but there is a real possibility they just leave things as is).

 

[Shoulder]

PS6 and the next Xbox have little chance of still supporting physical games. SNG is likely the last console to support it. Legacy content will mostly be preserved through the subscription services. Profits are substantially better with digital and consumers are gravitating towards it because it's convenient. I can emphasize with people who love physical, but it's nearing the end and no amount of bitching from a vocal minority is going to change that.

We can only hope. It feels like the entire industry is expecting an all digital future by the end of the decade.

The way I see it, all digital is inevitable, like what @Goodtwin said.

In many ways, it’s already happened. Look at the PC market. All digital, and I’m sure there were some enthusiasts, and other collector types who wanted to cherish and save those boxes the games came in. But it’s ultimately a niche market. An all digital future in the console industry is possible, and it won’t appear as dire as we might think provided we OWN the games we buy. Despite how popular Steam is, it’s still tied to DRM, which is where services like GOG are great especially for much older games. Sure, Valve’s intention is once you buy the game on Steam, it’s yours forever, though at anytime, that could be altered, especially if and when one day Gabe Newell leaves, or dies.

The biggest hurdle I think comes down to the service problem. Those games you buy ought to be available indefinitely, but contracts, licenses, and other agreements may or may not see it that way.

But going further than that, cartridges, and optical media do not last forever either, which is where digital comes in. Things such as Rom dumping are hugely important for the purposes of archiving, and while the legalese of it hasn’t necessarily been truly tested in court, it does bring up the question if we own the game, or any other piece of software. No one would sue you for taking a book, making notes on it, or outright removing pages from it would they? Of course not.

You'd still get all of those things. The Wii U was literally 3 Wii cpus fused together. It's BC mode was turning thr Wii U into an oversized Wii. That's not possible on Drake

Three Wii CPUs fused together, plus modified even further with additional cache, even in an asymmetric form between the cores themselves, plus clocked higher since it was on a much smaller node process than the Wii.

It’s a weird setup, and yet if fully realized may have been a decently powerful chip for developers. I recall back then that if it had 8 cores instead of three, it would’ve been as powerful overall as the AMD Jaguar, though maybe that was just Nintendo talk back in the day.

 

[NZZN]

If Switch 2 has a 1080p screen how is Nintendo going to avoid artifacting and other weird scaling issues when games run 720p in undocked mode?

Hopefully they implement a good spatial scaler like Lanczos or FSR 1 which has been featured in recent first party titles, bilinear is a worst case scenario they could use for minimal performance impact but produces output that's horribly blurry and aliased (see switch home screen and BC on most consoles).
There is a chance they could use Nvidia's NIS, which is supposedly optimised for Nvidia hardware but probably won't hold up to the more recent FSR.

edit: Here's a comparison of image scaling algorithms, FSR (AMD FidelityFX Super Resolution) is a fork of Lanczos with less processing time required for an almost identical output and an optional sharpening pass.

 

[Zedark]

If Switch 2 has a 1080p screen how is Nintendo going to avoid artifacting and other weird scaling issues when games run 720p in undocked mode?

(My) Baseline expectation - The same way they are dealing with current non-native games: just let them persist with these visual artifacts.

Lofty expectation - Perhaps they can employ a simple super-resolution algorithm/neural network on the tensor cores to upscale from an input resolution like 720p up to 1080p. In cases where native is not 720p, they could do simple upressing to 720p first if their algorithm doesn't handle variable input resolution. This would be purely a spatial algorithm, so it won't look as good as DLSS, but it would be better than naive local upscaling. The main question is how cheaply (rendering time-wise) such an algorithm can be employed: it obviously shouldn't be expensive since it must be bolted onto the rendering pipeline (as a final step) after the normal rendering phase. But I imagine a simple algorithm that beats naive upscaling doesn't have to be very expensive.

Perhaps they could additionally use optical flow information estimation to estimate some temporal information and enhance the super resolution that way.

 

[SoldierDelta]

they aren't even trying with these 'leaks'. of course 4k is going to be a feature that's what DLSS is there for.

I actually want to ask about some specifics around the Tensor Cores. I hear somewhere that the Switch 2 (if i remember correctly) will have 64 tensor cores. How many does the Switch need to utilise "Performance" settings for DLSS (Upscaling from 1080p to 4K for example)?

 

[Zedark]

Hopefully they implement a good spatial scaler like Lanczos or FSR 1 which has been featured in recent first party titles, bilinear is a worst case scenario they could use for minimal performance impact but produces output that's horribly blurry and aliased (see switch home screen and BC on most consoles).
There is a chance they could use Nvidia's NIS, which is supposedly optimised for Nvidia hardware but probably won't hold up to the more recent FSR.

I don't think FSR is an option since it needs to be slotted into the middle of the pipeline, effectively requiring you to open the rendering black box (which requires per-game patching). You would need something that applies super resolution to the final image of the native Switch game, after all the rendering post-processing steps have finished.

 

[NZZN]

I don't think FSR is an option since it needs to be slotted into the middle of the pipeline, effectively requiring you to open the rendering black box (which requires per-game patching). You would need something that applies super resolution to the final image of the native Switch game, after all the rendering post processing steps have finished.

FSR 2 functions like a combination of temporal anti-aliasing and a temporal upscaler so requires access to the rendering pipeline for motion vectors and pixel jittering.
FSR 1 just needs a finalised image to scale so can be used for image and video upscaling, for example.
MPV video player with FSR 1 scaling to stream my PSVita to my PC screen produced pretty impressive results for 960x544 > 1920x1080 so I can see it being closer to a native image for 720 > 1080 on a (relatively) small screen in the case of NG.

 

[Dekuman]

No leaks whatsoever yet. For the meantime, let's enjoy some reddit bullshit. It's a speculation thread after all.

* Hidden text: cannot be quoted. *

 

[Deleted member 887]

There don’t have to be scaling artifacts at all. A 720p “slice” of the screen would be 5.2 inches, basically the size of the Switch Lite.

Do I think Nintendo will do that? No, but there is an easy, artifact-free option

 

[ertaboy356b]

If Switch 2 has a 1080p screen how is Nintendo going to avoid artifacting and other weird scaling issues when games run 720p in undocked mode?

Maybe games can run in docked mode settings since the system will be very capable then

* Hidden text: cannot be quoted. *

in case people missed it since it's been deleted.

 

[ILikeFeet]

I actually want to ask about some specifics around the Tensor Cores. I hear somewhere that the Switch 2 (if i remember correctly) will have 64 tensor cores. How many does the Switch need to utilise "Performance" settings for DLSS (Upscaling from 1080p to 4K for example)?

Any number can do it. It's a matter of how fast it can. Folks have made estimations of up to 8ms to upscale to 4K. And given 60fps is 16.67ms, take that as you will

 

[Skittzo]

Actually, playing 720p BC titles on a 1080p screen could be a perfect use case for Nvidia's AI upscaler, right? That's applied to the video stream itself so it can be done on a system level.

 

[SoldierDelta]

Any number can do it. It's a matter of how fast it can. Folks have made estimations of up to 8ms to upscale to 4K. And given 60fps is 16.67ms, take that as you will

Can't wait for the Switch 2 to be able to run at 120 fps lmfaoo.... yeah a man can dream. I said this under the assumption that 120fps is 8.33ms, if i'm wrong pls correct/shoot me with a missile
Anyway, that's a really cool thing to hear about. That removes a lot of the concerns that the base switch had when it came to very mixed resolutions across many games.

One last thing I want to check, another thing I heard was that the Switch 2 has RT cores (I think 12?). Outside of ray-tracing, what could these cores be used for?

 

[Phenom08]

I see your point, but you make it seem like this thing - Nintendo’s undivided attention - is their key to success, and I just think there’s more to that. The Switch worked out very well, but you can’t do that over and over again and they have to get many things right to be successful. Not that they couldn’t, but I don’t think it’s gonna be easy either. Also, while they consolidated their resources, you can’t argue that development took them much longer over time. We’re not gonna get games like Majoras Mask made in just a year.

It is lol, Nintendo's consumers want to consume just as badly as Sony's or Microsoft's. Nintendo's consumers are still mostly gamers, those gamers are exactly the people that will buy Nintendo's hardware regardless of what it is because now there is no other choice. Doing it over and over is exactly what we see out of the other platforms. Nintendo obviously also provides other types of games to reach non traditional consumers, those people may not buy the next platform but that's certainly not the majority. Switch was an underpowered home console, over priced handheld, no 3rd party support at launch, tiny buttons, used a smart phone app to voice chat, lacks a messaging system, uses friendcodes, and has faulty controls. It has easily outsold the PS4 (Sonys most profitable and second highest selling home console) and may even become the highest selling system in history(beating PS2), you are kidding yourself to think that isn't on the back of Nintendo's ip.

 

[kvetcha]

Actually, playing 720p BC titles on a 1080p screen could be a perfect use case for Nvidia's AI upscaler, right? That's applied to the video stream itself so it can be done on a system level.

Kinda depends on the latency, I would imagine.

 

[ILikeFeet]

Can't wait for the Switch 2 to be able to run at 120 fps lmfaoo.... yeah a man can dream. I said this under the assumption that 120fps is 8.33ms, if i'm wrong pls correct/shoot me with a missile
Anyway, that's a really cool thing to hear about. That removes a lot of the concerns that the base switch had when it came to very mixed resolutions across many games.

One last thing I want to check, another thing I heard was that the Switch 2 has RT cores (I think 12?). Outside of ray-tracing, what could these cores be used for?

That was more of a mark against DLSS on Drake, actually. There's a fixed cpu cost that eats into that frame time, leaving you with less than that 16.67ms for gpu work. Essentially, 4k/60 will be out of the question for many 60fps games unless Nintendo and Nvidia make an even faster solution.

There's not much else for RT cores to do. It's designed for RT tasks, but luckily that's much wider range than people think

 

[Skittzo]

Kinda depends on the latency, I would imagine.

The latency probably comes from it being applied outside of the system I would imagine, when it's used now. If you can run it through the algorithm as close as possible to the video output I would think it could work without much (or any) latency, which they could do just as part of the BC overlay.

 

[kvetcha]

The latency probably comes from it being applied outside of the system I would imagine, when it's used now. If you can run it through the algorithm as close as possible to the video output I would think it could work without much (or any) latency, which they could do just as part of the BC overlay.

I defer to you! I haven't the slightest.

 

[SoldierDelta]

That was more of a mark against DLSS on Drake, actually. There's a fixed cpu cost that eats into that frame time, leaving you with less than that 16.67ms for gpu work. Essentially, 4k/60 will be out of the question for many 60fps games unless Nintendo and Nvidia make an even faster solution.

There's not much else for RT cores to do. It's designed for RT tasks, but luckily that's much wider range than people think

Ah, right sorry I kinda misinterpreted that.
Still, in theory, that shouldn't be too big of a problem. Even if the Switch 2 is a 1440p console, that's still the type of TV that most people have afaik. Additionally a lot of less demanding games can still make that 4K "ideal" so I'm not complaining.

 

[the_marmolade]

 

[TomNookYankees]

 

[Concernt]

* Hidden text: cannot be quoted. *

 

[JoshuaJSlone]

* Hidden text: cannot be quoted. *

That was before digital was a thing. There’s no point in making multiple slots now.

Unless digital can use transporter technology to convert old physical cards into digital form, there's still a point.

If Switch 2 has a 1080p screen how is Nintendo going to avoid artifacting and other weird scaling issues when games run 720p in undocked mode?

It's just not a big deal. If you'd notice a difference at all, it would be way less than something like DS on 3DS.

Actually, playing 720p BC titles on a 1080p screen could be a perfect use case for Nvidia's AI upscaler, right? That's applied to the video stream itself so it can be done on a system level.

The one they've recently used for YouTube and such? I don't think it's well suited vs things more like FSR1. A lot of what it does is cleaning up video compression artifacts, which is irrelevant when working with totally clean newly rendered game images.

 

[Concernt]

That was more of a mark against DLSS on Drake, actually. There's a fixed cpu cost that eats into that frame time, leaving you with less than that 16.67ms for gpu work. Essentially, 4k/60 will be out of the question for many 60fps games unless Nintendo and Nvidia make an even faster solution.

There's not much else for RT cores to do. It's designed for RT tasks, but luckily that's much wider range than people think

I think 4K60 will likely be as to NG Switch as 1080p60 is to Switch 1, at least as regards native rendering or DLSS. However I wouldn't discount games pulling additional tricks, like 1440pDLSS then 4K using a spacial upscaler.

If a dev WANTS 4K60, it can do it, that's something to keep in mind.

 

[chocolate_supra]

you can see how making DE was basically a stealth partial asset farm to help out 3 and FR's development.

I'm so glad I'm not the only person who thought this

* Hidden text: cannot be quoted. *

I actually quite like that

 

[Deleted member 887]

* Hidden text: cannot be quoted. *

The Wii was less powerful than the Xbox released 5 years earlier.

The Switch was modestly more powerful than the 360 and was a handheld.

Compared to the PS4 what is the “minimum” amount of power before the next handheld can be successful?