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

[ReddDreadtheLead]

what was the GB/s bandwidth ratio for PS4 and PS4 pro again?

Switch was definitely punching above its weight too for multiplatforms. It's too bad we'll never know how much 50 GB/s could have helped it. I think it was Z0mbie who said maxwell/turing graphics cards had 50 GB/s VRAM, and they were going toe to toe with Xbone/PS4

PS4 ratio should be around 95GB/s, while PS4 Pro should be ~52GB/s.


All this means is that relative to the FP32, PS4 had that memory bandwidth sufficient to support the system per second per TFLOP, it still delivered 176GB/s, but per TFLOP and for the GPU, we are only looking at the ratio. Does not mean that the switch NeXt will not be memory bandwidth starved.


This does not account for the fact that the PS4 CPU and PS4 Pro CPU use about 20GB/s if I remember right from a slide… but anyway, accounting for that, PS4 is ~84-85GB/s and PS4 Pro is ~47GB/s.


I’ll edit this post when I find that slide….


Actually, I’d like @oldpuck ’s opinion on this idea: measuring the memory bandwidth not with TFLOP, but number of SM/WGPs that a system/card has and comparing it there.

 

[Machrider]

VR can just be an optional feature for Switch 2, it doesn't have to be the whole enchilada.

For starters, I can see it being something that Nintendo recommends only using for 45-60 minutes per play session tops.

But that would still be plenty to enjoy like say a VR mode in Mario Kart Next or Pilotwings or Metroid Prime 4 or the next Mario platformer.

I think they are intrigued by it gameplay wise because if you really look at Labo VR ... Nintendo crammed a shit ton of content into it, likely because the VR idea fired up a ton of ideas in their development teams and when Nintendo hits on something like that, they tend to eventually iterate on it.

I don't think Labo VR will be the last time Nintendo does something with VR.

 

[ReddDreadtheLead]

Found it:

Actually I feel like this deserves a separate post…

This was clear, but not the “140GB/s” part.

 

[AoiHana]

Again and again, the chances of Nintendo succeeding in the VR market are slim to none.

Even with strong IPs like Mario, Pokemon, and Animal Crossing, they could not succeed in the mobile market (Pokemon GO is Niantic).

This will be the same for VR.

I wouldn’t go as far as to say they’ve failed in the mobile market. Pokémon Go (yes, technically Niantic) is the 6th-highest grossing mobile game of all time, and Fire Emblem Heroes has made over a billion dollars so far. They might not be pulling Genshin Impact or Candy Crush amounts of money, but by all accounts they’re making a killing. Even their less successful EPD efforts have still topped $100 million, for the most part. For a company that’s not primarily focused on mobile, they ain’t doing too badly at all. It’s just that traditional game development turned out more lucrative in the long run. (And thank god for that; can you imagine Nintendo going the way of Konami?)

I do sort of agree with your point on VR. The way I see it is if they changed their development philosophy to emphasize graphics and processing power, I’m sure they’d at the very least hold their own. For better or for worse, though, I don’t think they’re interested in taking the steps to do that at the moment. Insofar as their chances of success are “slim to none”, it’s because of their attitudes to game development rather than a lack of talent or skill.

 

[Terrell]

I said this elsewhere, so I'll happily repeat myself here.

It's important to recall that, at the start of the video game console business, the Atari 2600 alone managed to sell 10 million units in a little over 5 years, in an environment where people did not even know a video game at home was even possible. VR, with all of its advantages of already being part of humanity's collective imagination, failed to sell even as well as that collectively until Meta basically staked its entire company on the idea and poured billions into the effort.

Sometimes, the juice isn't worth the squeeze.

 

[Concernt]

I wouldn’t go as far as to say they’ve failed in the mobile market. Pokémon Go (yes, technically Niantic) is the 6th-highest grossing mobile game of all time, and Fire Emblem Heroes has made over a billion dollars so far. They might not be pulling Genshin Impact or Candy Crush amounts of money, but by all accounts they’re making a killing. Even their less successful EPD efforts have still topped $100 million, for the most part. For a company that’s not primarily focused on mobile, they ain’t doing too badly at all. It’s just that traditional game development turned out more lucrative in the long run. (And thank god for that; can you imagine Nintendo going the way of Konami?)

I do sort of agree with your point on VR. The way I see it is if they changed their development philosophy to emphasize graphics and processing power, I’m sure they’d at the very least hold their own. For better or for worse, though, I don’t think they’re interested in taking the steps to do that at the moment. Insofar as their chances of success are “slim to none”, it’s because of their attitudes to game development rather than a lack of talent or skill.

A good mobile showing from EPD can gross 100 million.

A good Switch showing from EPD can gross over a billion.

While I'm sure their mobile games don't take as much resources as a full scale Switch game, I'm sure it's a lot more than just a tenth!

My 2 cents on Nintendo's place in the fidelity race is that... If Drake launches this year, Nintendo is back in the running for having the best looking game on the market. I know I've worn this to death by now but it's true: it's not just a Nintendo console, it's the Nvidia console, and Nvidia has EVERY reason to make the console as powerful as they possibly can for the formfactor. Which is, as far as I can tell, exactly what they did with Nintendo Switch originally, and going forward, T239.

T239, if it's on 4N, would be a HUGE technical flex over AMD and Intel. Nvidia can get RT, DLSS, even 4K out of a 9W power budget and do everything except 4K on just 4W.

I continue to suspect 4N for a variety of reasons. Especially those power consumption tests, wowsa! 4K gaming at 9W of power consumption appears to genuinely be their aim with this thing!

I only hope it launches this year, as much as I have no reason to suspect it won't, I feel uneasy. That empty second half can't be anything BUT a new console launch. It just doesn't make sense otherwise.

 

[majimazio]

I wouldn’t go as far as to say they’ve failed in the mobile market. Pokémon Go (yes, technically Niantic) is the 6th-highest grossing mobile game of all time, and Fire Emblem Heroes has made over a billion dollars so far. They might not be pulling Genshin Impact or Candy Crush amounts of money, but by all accounts they’re making a killing. Even their less successful EPD efforts have still topped $100 million, for the most part. For a company that’s not primarily focused on mobile, they ain’t doing too badly at all. It’s just that traditional game development turned out more lucrative in the long run. (And thank god for that; can you imagine Nintendo going the way of Konami?)

I had forgotten about Fire Emblem. It's a miraculous success story.

However, killing Dr. Mario and Dragalia Lost by not expanding them to Switch at all is really unimpressive.

The Dr. Mario series that has continued on successive Nintendo consoles has ceased to exist, and the possibility that Dragalia could have been a JRPG brand like Xenoblade cannot be dismissed.

 

[Herb Alpert]

A good mobile showing from EPD can gross 100 million.

A good Switch showing from EPD can gross over a billion.

While I'm sure their mobile games don't take as much resources as a full scale Switch game, I'm sure it's a lot more than just a tenth!

My 2 cents on Nintendo's place in the fidelity race is that... If Drake launches this year, Nintendo is back in the running for having the best looking game on the market. I know I've worn this to death by now but it's true: it's not just a Nintendo console, it's the Nvidia console, and Nvidia has EVERY reason to make the console as powerful as they possibly can for the formfactor. Which is, as far as I can tell, exactly what they did with Nintendo Switch originally, and going forward, T239.

T239, if it's on 4N, would be a HUGE technical flex over AMD and Intel. Nvidia can get RT, DLSS, even 4K out of a 9W power budget and do everything except 4K on just 4W.

I continue to suspect 4N for a variety of reasons. Especially those power consumption tests, wowsa! 4K gaming at 9W of power consumption appears to genuinely be their aim with this thing!

I only hope it launches this year, as much as I have no reason to suspect it won't, I feel uneasy. That empty second half can't be anything BUT a new console launch. It just doesn't make sense otherwise.

Each and every year, in March, we know almost nothing about second half of the year...

 

[Concernt]

Each and every year, in March, we know almost nothing about second half of the year...

That's... Not true? Each and every year, especially after the first Direct, we know something about the second half of the year.

Knowing absolutely nothing IS exceptional, and it is dishonest to say that it isn't.

 

[ertaboy356b]

I had forgotten about Fire Emblem. It's a miraculous success story.

However, killing Dr. Mario and Dragalia Lost by not expanding them to Switch at all is really unimpressive.

The Dr. Mario series that has continued on successive Nintendo consoles has ceased to exist, and the possibility that Dragalia could have been a JRPG brand like Xenoblade cannot be dismissed.

They can make Dr. Mario as a side game to a new Super Mario game but as a standalone? I don't think it's worth it.

 

[Herb Alpert]

That's... Not true? Each and every year, especially after the first Direct, we know something about the second half of the year.

Knowing absolutely nothing IS exceptional, and it is dishonest to say that it isn't.

If they release their hardware next year, we can expect a lighter line up for the H2 2023, then it makes sense to not disclose to much in advance.

I don't know, you may be right, and I'd like you to be !
Anyway, investors meeting early may will be very telling, and we will be able to have a more precise idea. If they release New hardware during next fiscal year, they won't hide it to investors.

We should know soon enough !

 

[Terrell]

They can make Dr. Mario as a side game to a new Super Mario game but as a standalone? I don't think it's worth it.

Dr. Mario 99

There. I made it worth it.

 

[Mr Swine]

Although not directly related to Nintendo, I find this interesting.

Google (at 12:28) mentioned FSR 2 requires extra input information, which is why FSR 2's not usually used in a mobile environment.

Well it makes sense since FSR2 requires more “juice” to function good. Old low end GPU’s don’t benefit that much from FSR2 despite managing to get a 20% boost. It goes from 20fps to 24 which is not enough. Switch has Tensor cores which will take care of that

what was the GB/s bandwidth ratio for PS4 and PS4 pro again?

Switch was definitely punching above its weight too for multiplatforms. It's too bad we'll never know how much 50 GB/s could have helped it. I think it was Z0mbie who said maxwell/turing graphics cards had 50 GB/s VRAM, and they were going toe to toe with Xbone/PS4

Nvidia is way ahead of AMD when it comes to managing bandwidth. It’s going to be interesting to see Drake vs PS4 in multiplayer games like Resident Evil 4 Remake, Dying Light 2 and more despite having 65GB/s less bandwidth.

Also I wonder if RE4R will have certain settings from the Xbox Series S (shadows and lightning), X (Textures) and PS4 (Foliage and resolution) while having a stable 30fps

 

[Z0m3le]

What am I missing here, I didn’t know the Pokémon dev gave spec numbers?

They didn't. These are the only two posts that were made:


So I have no idea where the claim about memory speed came from.

I mixed up the 4chan leakers, this is the post I was referencing, I just had misremembered that it was pokemon, it's not that leaker. I just read Reddd's post about Ampere and CPU bandwidth, and that all popped into my head, so I decided to run the numbers and they fit. I am not backing this leak as creditable, but we can now see from my post, that 68GB/s and 102GB/s in portable and docked bandwidth speeds, are viable for Drake without causing performance issues. I also think the post is probably right about "fast" emmc, 400MB/s is probably enough for Nintendo's internal developers, and would be cheaper than UFS, especially in power consumption.

 

[ReddDreadtheLead]

Well it makes sense since FSR2 requires more “juice” to function good. Old low end GPU’s don’t benefit that much from FSR2 despite managing to get a 20% boost. It goes from 20fps to 24 which is not enough. Switch has Tensor cores which will take care of that

The GPU used is the Google Pixel 7 in set-fixed performance-mode enabled and at a fixed frequency.

 

[Dakhil]

Switch has Tensor cores which will take care of that

I think you meant to say Nintendo's new hardware equipped with Drake?
The GPU of the Tegra X1 inside the Nintendo Switch doesn't have Tensor cores.

 

[Simba1]

Huh? Series s is a 1080p to 1440p machine at best.

I have Series S, and I can certainly say that Series S is more than capable running plenty of games above 1440p if we dont talk about huge next gen AAA games and we talk about A-AA and 2D games, for instance Ori and the Will of the Wisps runs at 4K and 60FPS.

It's more thing that MS positioning it like 1080p-1440p console because Series X thats much more expensive is positing like 4K console, so they dont really pushing Series S above 1440p, and not about Series S is not capable running plenty of games at resolution above 1440p.

 

[Simba1]

IMO 1080p screen is most likely scenario for Switch 2 at this point.

What do you think?

 

[Dakhil]

IMO 1080p screen is most likely scenario for Switch 2 at this point.

What do you think?

Depends on if Nintendo wants to support VRR in both handheld mode and TV mode in my opinion.

 

[ILikeFeet]

I have Series S, and I can certainly say that Series S is more than capable running plenty of games above 1440p if we dont talk about huge next gen AAA games and we talk about A-AA and 2D games, for instance Ori and the Will of the Wisps runs at 4K and 60FPS.

It's more thing that MS positioning it like 1080p-1440p console because Series X thats much more expensive is positing like 4K console, so they dont really pushing Series S above 1440p, and not about Series S is not capable running plenty of games at resolution above 1440p.

They aren't holding things back due to positioning. If series s cam do 1440p or higher, the series X will just go higher. These games aren't the norm though. The games that would make up most purchases will struggle to hit 1080p. Some are struggling to be feature comparable to series x even

 

[Simba1]

They aren't holding things back due to positioning. If series s cam do 1440p or higher, the series X will just go higher. These games aren't the norm though. The games that would make up most purchases will struggle to hit 1080p. Some are struggling to be feature comparable to series x even

You cant know that, they could easily not even try achieving above 1440p resolution and saying partners to focus on resolution up to 1440p.
Like I wrote, I dont talk about big AAA games, you also have plenty of A-AA, 2D and various Indie games.

 

[Superizer]

VR faces two problems

1. People only want genuinely great VR that has no graphical issues that take them out of the experience or cause the graphics to look much worse
2. People will probably only pay up to $100 for VR

Problem 1 will probably be solved by like 2026

It will take another 8 or so years after that until VR headsets good enough to fulfill 1 are cheap enough to fulfill 2.

Devs have also done so little work on VR that like 99.9% of developers have no idea how VR games do or should work and that will also take a long time to get right.

Don't expect VR to be popular before 2035.

I never reply to something in this thread, because I have hardly any knowledge about the stuff discussed in this thread, but I want to add something to this.

A couple of months ago I had a chance to wear an AR headset for the first time, because a colleague of mine does research on using AR and he needed some people to complete some simple experiments using AR. I know that AR is not the same as VR but bear with me.
I wore the AR headset for around 10 minutes and I was really blown away by it. I did not expect it to be this good at all, because I always thought of AR and VR as some weird gimmick (just like the 3D technology for the 3DS). My line of thinking was along the lines of: "Why wear VR when you can just play on a TV screen? It has worked for a long time and is still great, why change it?"

I honestly think that a lot of people, at least many casual gamers (like me), have never worn a VR (or AR) headset and don't see much benefit in it. So I would add a third problem: getting people to try VR so that they know what they're missing out on. In my opinion it's hard to sell VR to the masses without having them try it.

 

[ILikeFeet]

You cant know that, they could easily not even try achieving above 1440p resolution and saying partners to focus on resolution up to 1440p.
Like I wrote, I dont talk about big AAA games, you also have plenty of A-AA, 2D and various Indie games.

The Series S being limiting isn't new news. The memory setup is fucking the system. Even with smaller games, if they could break through 1440p, they would if they could. Microsoft imposes no limits here. The Touryst runs at 4k on the series s. But look who's doing that, because not many small devs are Shinen

 

[Bosintang]

I'd agree if the JRPG library was more than 3 games.

I don't play jrpg and I have a bad image of those games, so I'm extremely biased in that regard.

IMO 1080p screen is most likely scenario for Switch 2 at this point.

What do you think?

I used to be firmly in the 720p camp. And I still think that we don't really need much more and I guess it can help saving battery life. But recently, my tea leaves have somehow started to tell me that 1080p was a likely outcome, unless mobile mode is significantly less powerful than docked mode. So I'm torn.
Whichever happens, I'll be happy. It's not an important factor to me.

 

[Cybergatuno]

IMO 1080p screen is most likely scenario for Switch 2 at this point.

What do you think?

I'm still on the fence about that.
Are we expecting a bigger screen than the OLED model?

The way I hold my Switch, I don't think I could tell the difference between 720p or 1080p for the same screen size. 1080p would come at the cost of battery and thermals to render and display more pixels. I think I'd rather get 720p and better framerate and/or battery.
If games consistently hit 720p, it's already quite an improvement. Even the Steam Deck chose 800p.

Phones are not comparable. We hold them a lot closer to our faces and they're optimized for comfort in short bursts, not sustained performance.

On the other hand, if it's 720p, there may be too wide of a gap between handheld and docked.

 

[Tentacle-tropes]

A good mobile showing from EPD can gross 100 million.

A good Switch showing from EPD can gross over a billion.

A good mobile showing from Nintendo can potentially earn way more then 100mil & reach the bil+ level. The biggest issues for Nintendo is that they don’t like the business model & they really didn’t take it that seriously. A long running successful mobile game would probably be one of their most expensive projects ever, perhaps even the most.

 

[Intoxicate]

IMO 1080p screen is most likely scenario for Switch 2 at this point.

What do you think?

Why is it the most likely scenario, I don’t see benefits for handheld play

I think no.

 

[KeijiKG]

VR faces two problems

1. People only want genuinely great VR that has no graphical issues that take them out of the experience or cause the graphics to look much worse
2. People will probably only pay up to $100 for VR

Problem 1 will probably be solved by like 2026

It will take another 8 or so years after that until VR headsets good enough to fulfill 1 are cheap enough to fulfill 2.

Devs have also done so little work on VR that like 99.9% of developers have no idea how VR games do or should work and that will also take a long time to get right.

Don't expect VR to be popular before 2035.

Affordability is always the best way to break barrier of entry and I'm pretty sure a Meta Quest VR runs on guts Switch 2 can match.
Even then, yeah. It's still more of a novelty at this stage or at least glorified 3D glasses support with headtracking. I can't see Nintendo considering VR unless they had some leftover Wii ideas that never left the Blackboard.

 

[Hermii]

IMO 1080p screen is most likely scenario for Switch 2 at this point.

What do you think?

Im 50/50 honestly. I also think an inbetween res like the SD is a possibility,

 

[Dekuman]

If they are disabling DLSS for portable 720p screen makes sense. They can run games native to that resolution.

Native 720p also looks great In a 6 to 7 inch screen so 1080 feels really wasteful. For a similar performance profile they can push more 720p native 60 fps games over 30 fps.

Fewer pixels to power also means a slightly less power hungry screen

 

[Deleted member 645]

I mixed up the 4chan leakers, this is the post I was referencing, I just had misremembered that it was pokemon, it's not that leaker. I just read Reddd's post about Ampere and CPU bandwidth, and that all popped into my head, so I decided to run the numbers and they fit. I am not backing this leak as creditable, but we can now see from my post, that 68GB/s and 102GB/s in portable and docked bandwidth speeds, are viable for Drake without causing performance issues. I also think the post is probably right about "fast" emmc, 400MB/s is probably enough for Nintendo's internal developers, and would be cheaper than UFS, especially in power consumption.

Thanks for clarifying that. I really hope Nintendo leverages their partnership with Nvidia. Switch 2 should be an easy swing and home run. I hope they don’t mess it up by attempting to be so outside of the box and different that alienate gamers and developers.

 

[kvetcha]

If they are disabling DLSS for portable 720p screen makes sense. They can run games native to that resolution.

Native 720p also looks great In a 6 to 7 inch screen so 1080 feels really wasteful. For a similar performance profile they can push more 720p native 60 fps games over 30 fps.

Fewer pixels to power also means a slightly less power hungry screen

Yeah, 720p to 800p feels like the sweet spot. Even if you had the bandwidth for 1080p you'd probably get a better looking picture at the lower resolution by just supersampling.

 

[Concernt]

If they are disabling DLSS for portable 720p screen makes sense. They can run games native to that resolution.

Native 720p also looks great In a 6 to 7 inch screen so 1080 feels really wasteful. For a similar performance profile they can push more 720p native 60 fps games over 30 fps.

Fewer pixels to power also means a slightly less power hungry screen

True, but we know they're not disabling DLSS hardware in Handheld and Tabletop mode.

Between that and the leaked power consumption tests targeting 1080p at 4W, I'm on team 1080p.

 

[kvetcha]

True, but we know they're not disabling DLSS hardware in Handheld and Tabletop mode.

Between that and the leaked power consumption tests targeting 1080p at 4W, I'm on team 1080p.

I don't think we know anything regarding DLSS usage.

 

[Concernt]

I don't think we know anything regarding DLSS usage.

Oh yes we do! NVN2 leaked, remember? It showed power consumption tests using DLSS with just 4W for the GPU, which is what handheld mode consumes. It also showed there's no way for developers to actually deactivate the DLSS hardware, Tensor Cores and OFA, depending on mode, so whether they use them or not they'll always be available.

 

[Deleted member 887]

Actually, I’d like @oldpuck ’s opinion on this idea: measuring the memory bandwidth not with TFLOP, but number of SM/WGPs that a system/card has and comparing it there.

Not a bad thought. My gut says that TFLOPS is still better - the idea is that, as the power of the GPU goes up, the more data it needs to touch over the course of a frame. On the other hand, the SMs are the thing you’re actually feeding, so there is more of a 1:1 relationship there.

 

[kvetcha]

Oh yes we do! NVN2 leaked, remember? It showed power consumption tests using DLSS with just 4W for the GPU, which is what handheld mode consumes. It also showed there's no way for developers to actually deactivate the DLSS hardware, Tensor Cores and OFA, depending on mode, so whether they use them or not they'll always be available.

The testing suite isn't rock solid proof of anything based on all the discourse here, only ballparking.

And regardless of the power state of the hardware, we don't know anything about how Nintendo is going to actually utilize it. It could just as readily be used to apply DLAA on a 720p or 800p screen.

 

[Goodtwin]

I have had the theory that Nintendo would use the OLED screen for Switch Redacted. They have been sourcing the screen for a couple years now and should be able to negotiate a superior price point going forward due to the volume they will be purchasing. However, many of the big upgrades we talk about really surround docked play more so than portable play. Being able to market a 1080p screen rather than 720p is going to get come gamers attention who play primarily in handheld mode. The difference on a 7" screen is not night and day, but seeing as how people have been willing to may for a superior screen of the same resolution with the OLED, a 1080p screen could be highly marketable. If the performance for DLSS is indeed very capable even at a 4w power draw, and extra watt for the screen wouldn't be a big deal since the SOC wouldn't be pulling anymore power than OG Switch. I am not confident either way, but if they do stick with 720p I really hope its a carry over of the OLED screen. Developers could also use DLSS to render at higher than 720p and then down sample for superior image quality.

 

[Foltzie]

A good mobile showing from Nintendo can potentially earn way more then 100mil & reach the bil+ level. The biggest issues for Nintendo is that they don’t like the business model & they really didn’t take it that seriously. A long running successful mobile game would probably be one of their most expensive projects ever, perhaps even the most.

I think that the restraint of having one mobile title per IP over ~7 years is "taking it seriously".

 

[Hermii]

Oh yes we do! NVN2 leaked, remember? It showed power consumption tests using DLSS with just 4W for the GPU, which is what handheld mode consumes. It also showed there's no way for developers to actually deactivate the DLSS hardware, Tensor Cores and OFA, depending on mode, so whether they use them or not they'll always be available.

The testing suite isn't rock solid proof of anything based on all the discourse here, only ballparking.

And regardless of the power state of the hardware, we don't know anything about how Nintendo is going to actually utilize it. It could just as readily be used to apply DLAA on a 720p or 800p screen.

Yep. Its very possible those power consumption numbers are not showing what we want them to show. It lacks context.

 

[Concernt]

The testing suite isn't rock solid proof of anything based on all the discourse here, only ballparking.

And regardless of the power state of the hardware, we don't know anything about how Nintendo is going to actually utilize it. It could just as readily be used to apply DLAA on a 720p or 800p screen.

Of course we can't know anything for certain, but it seems rather obvious, no? There is no provision made for deactivating the DLSS hardware in Handheld mode. So, to optimise power draw you actually want to be using it, not disregarding it. That's an awful lot of hardware just to do anti-aliasing.

Balance of probability, DLSS happens in handheld mode.

 

[Cybergatuno]

Oh yes we do! NVN2 leaked, remember? It showed power consumption tests using DLSS with just 4W for the GPU, which is what handheld mode consumes. It also showed there's no way for developers to actually deactivate the DLSS hardware, Tensor Cores and OFA, depending on mode, so whether they use them or not they'll always be available.

NVN2 was a (now dated) work in progress.
A lot of it may have been prototypes to pitch to Nintendo.
A lot of the fine tuning (implementing SoC features enabling/disabling, if supported) would be some of the final steps.

Hardware may have been taped out, but clocks, features and, specially, software will be a moving target until release, even beyond.

I don't think we can draw any strong conclusion.

 

[Simba1]

The Series S being limiting isn't new news. The memory setup is fucking the system. Even with smaller games, if they could break through 1440p, they would if they could. Microsoft imposes no limits here. The Touryst runs at 4k on the series s. But look who's doing that, because not many small devs are Shinen

Ori Will Of The Wisps is much better example, one of best looking 2D games ever, runs at 4K and 60fps.

 

[kvetcha]

Balance of probability, DLSS happens in handheld mode.

Conveniently ignoring my DLAA assertion.

DLAA supersampling is likely to provide better image quality than upscaling to 1080p native. And a high-quality 720p screen is going to be cheaper to produce.

 

[Raccoon]

I think the strongest argument for a 1080p display is marketing. Unfortunately good enough is never good enough, and after seven or eight years people will want a complete upgrade.

 

[Simba1]

Why is it the most likely scenario, I don’t see benefits for handheld play

I think no.

Because from much stronger hardware we could have 1080p games in handheld mode.

 

[Dakhil]

It also showed there's no way for developers to actually deactivate the DLSS hardware, Tensor Cores and OFA, depending on mode, so whether they use them or not they'll always be available.

Nobody knows if the OFA on Drake's GPU is going to have any meaningful use. The only known fact is that Drake's GPU inherits the same OFA as Orin's GPU. But how the performance of the OFA on Drake's GPU performance compares to the performance of the OFA on Ada Lovelace GPUs is unknown.

Hardware may have been taped out, but clocks, features and, specially, software will be a moving target until release, even beyond.

Hardware features are locked in place once the SoC's taped out.

 

[kvetcha]

Because from much stronger hardware we could have 1080p games in handheld mode.

Yeah but at the screen sizes we're dealing with you'd be just as well off going 800p. The difference in pixel density is not going to be obvious unless you're playing at myopic viewing distances.

 

[Tentacle-tropes]

I think that the restraint of having one mobile title per IP over ~7 years is "taking it seriously".

It really isn’t. Especially with their stated goals for mobile & hesitance to use the business model in other mobile titles even till this day. This in conjunction with a seeming pull back really points to a strategy of appeasing shareholders & trying to buy time for Switch to generate money so they could exit.

 

[Raccoon]

I'd argue that deploying internal development resources to creating and supporting Mario Kart for phones absolutely indicates a serious commitment to that space