StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (New Staff Post, Please read)

[Hosermess]

Yeah I mean if someone invokes the word 3 times he has to answer him, obviously. He could not possibly just give a non-answer like "we continue to evaluate future hardware plans to satisfy out customers and grow our business" - since the guy said Next gen 3 times he had to give specific details on his next generation console. I'm sure he's kicking himself in his office right now thinking about how this investor outsmarted him like that.

I guess I don't have the right kind of eyes, cause I did not see any specific details about hardware in his answer.

Software, Nintendo Accounts and non-gaming channels. That's their plan "to maneuver to the next generation of hardware." What that next generation will be is still a mystery.

 

[Serif]

So, it won’t be Wii U to the Wii, more like 3DS to DS.

I also think it will be like GBA to GB. Same "brand" (Switch, Game Boy) to leverage its success, much more powerful internals, new play mechanics / gimmicks in the form of peripherals, enhanced backwards compatibility. But of course, with modern games being much more scalable, they can have an extended cross-gen period with enhanced Switch titles.

(Also, I'm never giving up my affinity for the 'Switch Advance' moniker. I love it.)

 

[Hosermess]

fuck

 

[chocolate_supra]

fuck

Honestly at this point I'm just like:

 

[Foltzie]

Honestly at this point I'm just like:

Which group complains about the marketing.

 

[ShadowFox08]

Honestly at this point I'm just like:

huge difference. Revisions use the same tech for the most part (notably cpu) and are natively backwards compatible, and don't usually have exclusive games. Successors are new everything.

 

[Deleted member 887]

re: the next generation "concern" answer

I want to point out that Furukawa is confirming something here that the thread has brought up a number of times. We can't entirely look to Nintendo's past behavior on revisions/replacements to predict it's future, because Nintendo's past hardware cycles were driven by having two distinct product lines.

In a world where Nintendo can no longer rely on the GameBoy/DS lines to act as a safety net when jumping TV console generations (and vice versa), Nintendo can't afford a failed console - yet. How they approach the revision/successor is going to be different than it ever was before because of that.

Nintendo is in good shape currently, but the "all eggs in the Switch basket" strategy has them walking a tightrope. Nintendo accounts, making Mario Kart a secret live service game, theme parks, movies and TV shows, even their mostly underwhelming mobile strategy is the net they're building underneath themselves in real time. We're going to see more and more movie and TV deals, because Nintendo needs that licensing money as a constant stream of hardware independent revenue, and because they need a way to make and preserve Nintendo fans that functions regardless of what game comes out.

This is why I've been able to believe this non-traditional "revcessor" approach, but even if that's not what Nintendo does, Furukawa linking the media strategy to their hardware strategy is a clear indicator that Nintendo knows it needs to untangle it's revenue from the boom/bust hardware cycle.

 

[Stinkmeiser45]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

 

[Skittzo]

huge difference. Revisions use the same tech for the most part (notably cpu) and are natively backwards compatible, and don't usually have exclusive games. Successors are new everything.

No, the correct answer is "marketing".

That's the only difference.

 

[Dark Cloud]

Time for me to Cherry pick Furukawa’s answers to fuel the narrative I want for the next system.

 

[Stinkmeiser45]

Time for me to Cherry pick Furukawa’s answers to fuel the narrative I want for the next system.

Or we could instead blindly trust some random dude on youtube with 0 track record.

 

[metonight]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

Those tools are constantly updated, even for the current model. Maybe someone left at Nvidia and now they need someone else? So it shoudn't mean much.

 

[Hosermess]

.

 

[ILikeFeet]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

there has been similar ads in the past so there's now way to judge timing with them

 

[Dakhil]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

That's not necessarily true since there's a similar job listing posted by Nvidia on LinkedIn on November 2021, which I posted on the OP.

 

[Tentacle-tropes]

Is this the part that so many outlets, tweets, and threads had initially shared as Nintendo being "concerned" about the next-gen transition?

Yes, sadly so many outlets, tweets, threads, & videos use the mistranslated title or wording way to often then never correct themselves.

 

[Deleted member 887]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development.

Tools engineer is different from API developer - this seems to mostly be a GPU profiler. The sort of thing I would expect to update regularly after device launch, and also the sort of thing that tends to lag behind everything else. A common complaint of game devs working on launch games is that the tools are bad/non-existent.

This was one of the things that Nintendo had to throw away when the left the Wii U. By all accounts the Wii SDK and tools were incredible at launch, because they were able to carry over years of investment in the GameCube architecture

 

[Skittzo]

Or we could instead blindly trust some random dude on youtube with 0 track record.

We're not the ones trying to create a narrative here, and certainly not based on anyone in particular so I'm not sure what you're talking about regarding YouTube.

You: "look at this answer, the phrasing suggests X"
Us: "nah I don't think it tells us anything"

That's all that's happening here.

 

[Raccoon]

Or we could instead blindly trust some random dude on youtube with 0 track record.

or we could both ignore the iffy rumors and not draw conclusions from innocuous statements of long-term strategy

 

[Falcatops]

I don’t get why we are spiraling again into a successor vs revision discussion or a ‘22vs’23vs’24 one.
Yes, Question #4 mentions next-gen hardware 2.5 times (2 direct and 1 indirect mention), but Furukawa’s Answer #4 can be broken down like this from the official translation:

We have already announced some software titles that will be released through next spring.
Unlike in the past, even though Nintendo Switch has gone through five years since its launch, there is still a rich lineup of new titles to be released.

Here he’s just stating facts.

The biggest reason for this is that, thanks to the smooth launch of Nintendo Switch itself, we have been able to concentrate our development resources on one platform.

He’s mentioning the benefits of consolidating their development teams, console+handheld vs hybrid and all that. We already know this.

On the other hand, looking back on past experiences of generational change such as the change from the Wii and Nintendo DS eras, we recognize that one of our tasks is ensuring the transition to future generations of hardware is as smooth as possible.

Here he realizes that in the past Nintendo has had problems moving to a new platform, specially from a successful one (Wii+DS). But he doesn’t specify if the goal of ”ensuring a smooth transition” is for 1) the benefit of the platform holder (they), 2) the benefit of devs in general (1st & 3rd party) or 3) our benefit as consumers. Maybe it’s for all 3 of them. Please note that here I’m talking about who benefits the most from the situation, not about the measures they will take to accomplish said transition.

To that end, we are focusing on building long-term relationships with our consumers (through Nintendo Accounts).

Wait, does this mean that in the previous point he was talking about option #3? Not necessarily. A long-term relantionship with consumers can be beneficial to all involved. For example, they could be looking to do a smooth transition for 3rd party devs selling their games by ensuring that people move to new hardware with their accounts and keep inside the ecosystem. We already knew about this strategy from a previous financial report infographic.

While continuing to release new Nintendo Switch software for consumers to enjoy, we aim to maintain relationships across hardware generations through services that utilize Nintendo Accounts and by providing opportunities for them to experience our IP through other non-gaming channels.

Mix of new and old. We knew about non-gaming channels: the theme park, movies, merchandise, etc.
New info (as far as I remember): when a new devices eventually releases, regardless of whether it is a revision or a successor, the current Nintendo Switch will keep getting new games, and the new device will have BC with the current Switch at least from the digital library. The answer does not discard physical BC.

And it works both ways. If Drake is a Pro then it’s obvious that it will play current games. And if it’s a Switch 2, then maybe we are getting an extended cross-gen period like with PS5 and Xbox Series.

So we are exactly where we started. A non-commiting PR answer.

 

[fwd-bwd]

Is this the part that so many outlets, tweets, and threads had initially shared as Nintendo being "concerned" about the next-gen transition?

Yes, sadly so many outlets, tweets, threads, & videos use the mistranslated title or wording way to often then never correct themselves.

When reading any seemingly significant/sensational claims, I always check the source that they cited. In this case, all of them originated from one (1) VGC article. VGC has a history of dodgy translations. I don't know how it could go from "one of our tasks" (official) to "a major concern" (VGC) without the mistranslation being deliberate. The internet destroyed the traditional business model of journalism, therefore many outlets resort to these clickbaits (and Facebook/Twitter algorithms further amplify misinformation for profits—"free speech" champions they are not). It's for the audience to be vigilant about what they read, and calls out the bad practice when it's particularly egregious.

 

[My Tulpa]

Separate handheld and console could work in theory maybe like 5-10 years down the line, especially if they plan to use that SCD technology to prop up the handheld's processing capability.

As long as they share the same software ecosystem I don't really see a problem, it's still extremely far removed from what they were doing before the Switch.

The next Nintendo gaming hardware could be a separate handheld and a home console sharing the same library.

Could we really see this concept and what would be the pros and cons?

The home console version could be sold for cheaper than a Series S and it would bring down the cost of the handheld as well since it wouldn’t need to be packaged with a dock and the internals would only need to contain lower clocked hardware.

You’d be able to share your games and saves across both home and handheld systems if you had both.

If they share the exact same games…why bother splitting the hardware then? Doesn’t make sense to me.

Nintendo isn’t going to make a tv only console with so much of a power differential over the portable because they wouldn’t design games around the expensive power model. They will always design and develop their games around the portable baseline and scale up.

So just keep the hardware combined, in that case.

As I said before, Switch tv only will be the next step. More power hungry. Current model to play on the go and tv model to play in 4 k the same cartridges. An hdd to download 4k textures.

You can upscale Nintendo games to your 4K tv just fine and still keep it a hybrid. No need for the home console only option if it’s just for that.

Original Switch was heavily undershipped, Nintendo didn't expect it to sell so much during launch, I think if they could they would ship a lot more units than they did in 2017 even with the stock issues we are having.

Yes, but even with a “normal” launch quarter period of selling/shipping ~4 million units or so…it still validates what the poster you were responding to was saying.

The chip/component shortage has no relation to whether new hardware launches or doesn’t.

So, that 21 million projection could easily include ~4 million of Drake sales in Q4 and expecting to sell ~17 million of the cheaper models for the FY.

 

[My Tulpa]

Ah yeah, that's kinda different from the typical "Nintendo gonna Nintendo". I tend to agree with Nate, Nintendo loves to experiment with their hardware and I have a hard time believing they'll never try something new that'll necessitate dropping the hybrid form factor, at least partially.

lol “Nintendo gonna Nintendo” is way nicer way to put it than what NateDrake said.

He basically said that Nintendo trying something different with their hardware directly after a success is “stupid”

 

[My Tulpa]

This smooth transition talk made me start believing in two things regarding the new hardware:

1. BC has to be a fact, there's no way Nintendo will make a new Switch without BC, but I believe that it will be digital-only and you'll somehow have to register your physical copy to get a digital one.
2. March 2023 seems to be the release date they'll choose because the other aspect you have to consider when you focus on "smooth transition" is sales, you can't wait for your sales to fall from a cliff to release new hardware, you gotta keep the momentum on your side. So the way I see it, they expect the new Switch to sell around 2 million in the current fiscal year and the old Switch to sell around 19 million, giving them the 21 million they forecasted.

That’s funny, this “smooth transition” talk just reinforces my opinion that whatever the revision they are releasing soon won’t be treated as a successor/gen break model.

It will be treated as a mid gen revision meant to elongate the Switch lifespan, nothing more.

And the talk of that next gen machine he was asked about was for whatever comes after, years from now.

This seems like such an unimportant statement but there's so much to unpack from the phrasing here
1) Nintendo is currently planning a transition to a new generation of hardware, he specifically says generation and not "revision" or "model", which means the idea of Nintendo moving to an iterative approach is now dead.
2) Because they are currently planning this transition, it probably isn't that far away. I would say this largely deconfirms a release later than 2024, at the absolute latest. Most likely though we see it next year.
3) With Nintendo planning a generational transition in the near future, a pro model is likely not happening. More likely is something similar to a PS5, where it will have a similar form factor and name and some games may be cross generation, but it will also have its own identity and exclusives.

You can say that I'm overthinking this, but I doubt it.

Ok, I’ll say it…you’re overthinking this.

The guy who asked him the question brought up “next generation hardware”…not Nintendo. Furukawa simply answered the guys question about how they navigate what’s coming after the Switch. You know, when it’s lifecycle is over…which Furukawa keeps saying will be unusually long…

They can just not answer? Or even just be more vague about it. Say something generic like "we always think about how to transition to new products."

That’s basically what he did, though.

 

[Falcatops]

For the more technical folks. A few pages back there was a link to a review/preview/whatever of AMD’s FSR 2.0 vs DLSS 2.0 with the ”DLSS killer” bait title.

Well my question goes along the way of: are there any comparisons of how much time it takes for FSR and DLSS to upsample a frame? Say, from 1080p to 4K. I know that DLSS on Ampere is faster than on Turing, because Turing has to proccess DLSS after the frame is done rendering while Ampere can do both tasks concurrently.

That review said that a FSR 2.0 4K frame is very similar in quality to a DLSS 2.0+ one. And it’s obvious that given enough time/money both solutions will reach a point of maturity where their results are mostly the same.
But FSR uses the regular shader cores on the GPU that could be used for regular rendering, while DLSS offloads that work to the AI specialized Tensor core and lets the Cuda cores do their thing.

So, is there any evidence that shows noticeable benefits when using the tensor cores? Because we can all agree that Drake will need all the shader power it can get.

 

[My Tulpa]

Honestly at this point I'm just like:

A successor console is meant to replace the previous console’s sales as soon as possible…both in terms of hardware sales/production and in terms of software development focus.

A revision console is meant to elongate the lifecycle of the current console ecosystem, selling alongside a multiple of value option models while sharing basically the same software development focus.

This is the only answer.

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

This reminds me all the times people show a big new game hiring new people and assuming that means the game is years away…and then it goes gold 6 months later lol

Same energy.

And it works both ways. If Drake is a Pro then it’s obvious that it will play current games. And if it’s a Switch 2, then maybe we are getting an extended cross-gen period like with PS5 and Xbox Series.

Except Microsoft and Sony stopped making ps4’s and Xbox one’s when ps5 and Xbox launched. (In some cases, production stopped before launch)

I seriously doubt if Nintendo releases a $400-$500 new model within the next 12 months, that they are going to discontinue the current models. I also seriously doubt Drake has any Nintendo exclusives on it during its launch year. Why would Nintendo ignore having a profile that targets the Lite/OLED profile? Cause they think OLED specs are holding back their 1st party games? lol no

 

[Falcatops]

Except Microsoft and Sony stopped making ps4’s and Xbox one’s when ps5 and Xbox launched. (In some cases, production stopped before launch)

Ok you’re not wrong because you’re stating facts. And I never said that Nintendo is going to stop or continue production.

I seriously doubt if Nintendo releases a $400-$500 new model within the next 12 months, that they are going to discontinue the current models.

I never said that.

I also seriously doubt Drake has any Nintendo exclusives on it during its launch year.

I never said that Drake will have exclusives.

Why would Nintendo ignore having a profile that targets the Lite/OLED profile?

No idea, I don’t make the calls at Nintendo.

Cause they think OLED specs are holding back their 1st party games? lol no

Who knows? Maybe they do, maybe they don’t.

All I said was that the only thing we can extract from Furukawa’s answer (quoted again below):

While continuing to release new Nintendo Switch software for consumers to enjoy, we aim to maintain relationships across hardware generations through services that utilize Nintendo Accounts

Is that their next device, whatever form it takes, will have BC to ”ensure a smooth transition”. And we kinda expected that already because they will keep using an Nvidia chip.

 

[Anatole]

For the more technical folks. A few pages back there was a link to a review/preview/whatever of AMD’s FSR 2.0 vs DLSS 2.0 with the ”DLSS killer” bait title.

Well my question goes along the way of: are there any comparisons of how much time it takes for FSR and DLSS to upsample a frame? Say, from 1080p to 4K. I know that DLSS on Ampere is faster than on Turing, because Turing has to proccess DLSS after the frame is done rendering while Ampere can do both tasks concurrently.

That review said that a FSR 2.0 4K frame is very similar in quality to a DLSS 2.0+ one. And it’s obvious that given enough time/money both solutions will reach a point of maturity where their results are mostly the same.
But FSR uses the regular shader cores on the GPU that could be used for regular rendering, while DLSS offloads that work to the AI specialized Tensor core and lets the Cuda cores do their thing.

So, is there any evidence that shows noticeable benefits when using the tensor cores? Because we can all agree that Drake will need all the shader power it can get.

AMD reports between 0.5 ms and 1.5 ms for different GPUs for FSR 2.0.

I’m not sure what exactly you mean by your last question, but the presentation slides for FSR 2.0, they say that “a lot of ALU cycles” are used to calculate the Lanczos weights, which upscales and filters the rendered samples.

DLSS instead uses a type of neural network called a convolutional autoencoder for filtering samples at that step, which in practice means running matrix multiplications on the tensor cores.

 

[My Tulpa]

Ok you’re not wrong because you’re stating facts. And I never said that Nintendo is going to stop or continue production.

I never said that.

I never said that Drake will have exclusives.

Sorry, I was just addressing the part where you were suggesting the possibility of Drake acting like a “Switch 2” the same way the ps5 and Xbox Series did.

I was listing the reasons why that wouldn’t happen or act anything like those systems.

 

[YolkFolk]

In my opinion the drake is Switch 2. Like others I feel I should be allowed to have this opinion without being shouted down by My Tulpa.

The jump in hardware over Switch 1 is a generational leap and then some especially when DLSS is taken into account. You don’t make a device with this much horse power only for it to be a used for increased resolution or the odd third party exclusive.

My take is that first party games will release on Switch 1 and Switch 2 for a couple of years before they’re Switch 2 exclusive. Third parties will make the transition at a faster rate with exclusives right out of the gate. I’d imagine games like Resident Evil Village and Elden Ring will appear within 12 months of launch.

Hey, I could be wrong but this is my take. Hopefully I won’t now face a post where they points are “taken apart” bit by bit.

 

[Skittzo]

In my opinion the drake is Switch 2. Like others I feel I should be allowed to have this opinion without being shouted down by My Tulpa.

The jump in hardware over Switch 1 is a generational leap and then some especially when DLSS is taken into account. You don’t make a device with this much horse power only for it to be a used for increased resolution or the odd third party exclusive.

My take is that first party games will release on Switch 1 and Switch 2 for a couple of years before they’re Switch 2 exclusive. Third parties will make the transition at a faster rate with exclusives right out of the gate. I’d imagine games like Resident Evil Village and Elden Ring will appear within 12 months of launch.

Hey, I could be wrong but this is my take. Hopefully I won’t now face a post where they points are “taken apart” bit by bit.

Nothing wrong with your opinion, but I personally disagree with the premise that it has to be one or the other, when instead I expect it to essentially be both.

 

[Deleted member 887]

Latest DF video on FSR 2.0 is very good, and FSR 2.0 looks great. Worth checking out for this thread because it (incidentally) makes a case for why DLSS is such a good win for Nintendo. To sum up

• DLSS performance is generally better, and likely to always be because tensor cores.
• FSR runs on any card - but that doesn't matter in the console space
• In comparison, current DLSS seems consistently better when working in performance mode.

 

[ReddDreadtheLead]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

Not at all, XBox and PS5 launched with unfinished developer kits , they got updated later on.

Incorporating things like FSR or something like PSVR2.

And with this not being the first job listing about this (there was one in like November and another before that), you can’t really draw any conclusion on timing with this.

 

[chocolate_supra]

In my opinion the drake is Switch 2. Like others I feel I should be allowed to have this opinion without being shouted down by My Tulpa.

The jump in hardware over Switch 1 is a generational leap and then some especially when DLSS is taken into account. You don’t make a device with this much horse power only for it to be a used for increased resolution or the odd third party exclusive.

My take is that first party games will release on Switch 1 and Switch 2 for a couple of years before they’re Switch 2 exclusive. Third parties will make the transition at a faster rate with exclusives right out of the gate. I’d imagine games like Resident Evil Village and Elden Ring will appear within 12 months of launch.

Hey, I could be wrong but this is my take. Hopefully I won’t now face a post where they points are “taken apart” bit by bit.

From what I've been seeing in all my lurking, and in my own personal opinion, I think what you're describing is largely the consensus with the only thing drawing out arguments being how Nintendo will market it. I don't think anyone is arguing that it won't be a generational leap in performance or that there won't be a long period of crossover support before Mariko is phased out and Drake is the new "base." The big question continues to be whether Nintendo comes out and calls this something akin to "Switch 2" or if they do what they've done in the past and consider their generational breaks as being a change-up in gimmick, thus considering even an outrageously more powerful Switch "still a Switch."

 

[ReddDreadtheLead]

By all accounts the Wii SDK and tools were incredible at launch, because they were able to carry over years of investment in the GameCube architecture

Actually, I’ve always heard from comments from devs that the Wii SDK was terrible. And looking at the ones for the other consoles, it seems like Nintendo has had a history of having a terrible SDK for developers. With the Switch being the first time it was very easy and accessible for developers to bring games towards the platform.

Even though the Wii was very similar to the GCN.

 

[manfromthedam]

Completely based on my own tastes, but if there's a single third party game that I suspect will make it over to Drake it'll be the new Need For Speed by Criterion. It's a late 2022 game, and EA put both Burnout Paradise and Hot Pursuit 2010 on Switch - seemingly arbitrarily. Not so much if you see it as preparing the market for a new NFS.

/fanfic

 

[LiC]

Nintendo Switch 2: Nvidia Hiring for Next-Gen Developers Console Tool

According to a recent job listing, Nvidia is hiring to develop tools for next-gen console gaming hinting at a possible Nintendo Switch 2

tech4gamers.com

A new hiring ad put out by nvidia for a game console developer tools engineer, presumably for Drake and specifically NVN2.
Interesting that NVN2 appears to still be under active development. While by no means a deconfirmation, I'd say it makes an early 2023 release less likely. It will take time for developers to port their games to the console, so if the development tools aren't largely finalized that's gonna be a little hard to do in time.

NVN1 and Switch devtool development has been continuous since 2014. Development on NVN2 and the new model's devtools will be continuous too. Hiring more people for the project -- if that's what this job listing is for -- doesn't mean anything with regards to how close it is to release, close or far.

Tools engineer is different from API developer - this seems to mostly be a GPU profiler. The sort of thing I would expect to update regularly after device launch, and also the sort of thing that tends to lag behind everything else. A common complaint of game devs working on launch games is that the tools are bad/non-existent.

This was one of the things that Nintendo had to throw away when the left the Wii U. By all accounts the Wii SDK and tools were incredible at launch, because they were able to carry over years of investment in the GameCube architecture

Indeed. There are several debugging and profiling tools in the NVN1 source, and one of them (a crash dump debugger) seems to have begun development as late as 2020.

 

[Falcatops]

I’m not sure what exactly you mean by your last question, but the presentation slides for FSR 2.0, they say that “a lot of ALU cycles” are used to calculate the Lanczos weights, which upscales and filters the rendered samples.

DLSS instead uses a type of neural network called a convolutional autoencoder for filtering samples at that step, which in practice means running matrix multiplications on the tensor cores.

Thanks for your answer. Sorry if the question is confusing, english is not my first language. Let me rephrase it again:

1) Considering the following facts:

• FSR runs on the “general shading cores” (ALU, CUDA, etc).
• DLSS runs on the “specialized cores” (Tensor).
• Both solutions look similar to the untrained eye.
• Graphics processing power is always limited. FSR has to be accounted for in the direct processing power budget available to a frame while DLSS does not.
• Silicon space is always limited, when you add Tensor cores to the SM you “lose” space that could go to adding more CUDA cores. When not using DLSS, you effectively lose functional silicon space.
• Not all games will use FSR or DLSS.

2) Has anyone done an analysis of how much time does it takes to upsample a frame using FSR vs DLSS on GPUs of equal power? (So no RTX 2060 vs RX 6800XT)

What I want to know: is it better to have and use Tensor cores for DLSS vs just adding more CUDA cores for general graphics power and the occasional FSR?

Is the difference significant in terms of framerate? As in “using DLSS on this GPU let us go from 35 to 60 fps at 4K” vs “using FSR on the same GPU lets us go from 35 to 52 fps at 4K”

Sorry, I was just addressing the part where you were suggesting the possibility of Drake acting like a “Switch 2” the same way the ps5 and Xbox Series did.

I was listing the reasons why that wouldn’t happen or act anything like those systems.

Ah don’t worry, I was just giving a general example. My main point in that post is that we can’t take Furukawa’s answer as a confirmation of anything.

---

Edit: This is the DF video mentioned by @oldpuck


Edit 2: thanks again for mentioning the video. I’m suscribed to DF but didn’t get a notification.
Just finished watching it and it’s just what I was asking for.

 

[My Tulpa]

In my opinion the drake is Switch 2. Like others I feel I should be allowed to have this opinion without being shouted down by My Tulpa.

The jump in hardware over Switch 1 is a generational leap and then some especially when DLSS is taken into account. You don’t make a device with this much horse power only for it to be a used for increased resolution or the odd third party exclusive.

My take is that first party games will release on Switch 1 and Switch 2 for a couple of years before they’re Switch 2 exclusive. Third parties will make the transition at a faster rate with exclusives right out of the gate. I’d imagine games like Resident Evil Village and Elden Ring will appear within 12 months of launch.

Hey, I could be wrong but this is my take. Hopefully I won’t now face a post where they points are “taken apart” bit by bit.

Sorry if my posting style is off putting, I really don’t mean to come off as yelling or aggressive. Just giving my firm opinion like others are.

Just curious, when do you think Drake will release then?

 

[ILikeFeet]

2) Has anyone done an analysis of how much time does it takes to upsample a frame using FSR vs DLSS on GPUs of equal power? (So no RTX 2060 vs RX 6800XT)

Digital Foundry did a comparison with 1 GPU (the 3080). they also find that "similar power" does not equal similar times as Ampere was always faster than RDNA2. so architectural differences are at play

 

[karmitt]

Latest DF video on FSR 2.0 is very good, and FSR 2.0 looks great. Worth checking out for this thread because it (incidentally) makes a case for why DLSS is such a good win for Nintendo. To sum up

• DLSS performance is generally better, and likely to always be because tensor cores.
• FSR runs on any card - but that doesn't matter in the console space
• In comparison, current DLSS seems consistently better when working in performance mode.

I was thinking the same thing, but it does kind of hurt prospective Switch revision ports a bit, no?

There was the belief that DLSS would be the special sauce to help keep the revision in arms-length of the home consoles. If FSR 2.X ends up being applied well on most future XSX titles for instance, it distances things once again.

 

[ArchedThunder]

DLSS is quite a bit cheaper and so far seems to be of a better quality than FSR 2.0. The cost of FSR 2.0 might see many devs just go with cheaper temporal upscaling.

 

[kimbo99]

All this talk of Drake when Kendricks album released today.

 

[ReddDreadtheLead]

All this talk of Drake when Kendricks album released today.

Talent has won today.

 

[Deleted member 887]

I was thinking the same thing, but it does kind of hurt prospective Switch revision ports a bit, no?

There was the belief that DLSS would be the special sauce to help keep the revision in arms-length of the home consoles. If FSR 2.X ends up being applied well on most future XSX titles for instance, it distances things once again.

Yes and no. FSR will help squeeze better and better images out of existing hardware, but if every game has to use a temporal upscaler, then the advantage goes to the console that hardware accelerates that upscaler over the console that has to spend general purpose power on it.

XSX can draw more electricity out of the wall and has more silicon, it will always win. But if tensor cores take up (totally random number) 20% of the New Switch's transistors, but FSR takes 30% of the XSXs to run, then the gap becomes much smaller than it looks on paper.

 

[ILikeFeet]

All this talk of Drake when Kendricks album released today.

I've been busy so I haven't had a chance to listen yet

I was thinking the same thing, but it does kind of hurt prospective Switch revision ports a bit, no?

There was the belief that DLSS would be the special sauce to help keep the revision in arms-length of the home consoles. If FSR 2.X ends up being applied well on most future XSX titles for instance, it distances things once again.

no one should have it in them that Drake could do PS5/Series X without cutback. I think one should expect cutbacks even compared to the Series S, which is largely Series X but at reduced resolution and sans RT usually

 

[Deleted member 887]

Actually, I’ve always heard from comments from devs that the Wii SDK was terrible. And looking at the ones for the other consoles, it seems like Nintendo has had a history of having a terrible SDK for developers. With the Switch being the first time it was very easy and accessible for developers to bring games towards the platform.

Even though the Wii was very similar to the GCN.

This is admittedly all second hand info - - my understanding was that the SDK was very mature. Lots of stuff - like documentation written in English - that doesn't usually come day 1 was present and correct. Getting a game up and running was apparently easy. Getting anywhere resembling sane performance out of it was not. The Switch's flat architecture means you never have to go around the SDK to get good performance, the weird Wii (or much worse, Wii U) design meant that going around it was a requirement.

 

[karmitt]

I've been busy so I haven't had a chance to listen yet


no one should have it in them that Drake could do PS5/Series X without cutback. I think one should expect cutbacks even compared to the Series S, which is largely Series X but at reduced resolution and sans RT usually

I never said “without cutbacks.” I said arms-length, which is only implying porting being viable. My concern was that any supposed gains from DLSS would be erased with the adoption FSR 2.0

 

[ILikeFeet]

I never said “without cutbacks.” I said arms-length, which is only implying porting being viable. My concern was that any supposed gains from DLSS would be erased with the adoption FSR 2.0

unless games blow out the cpu budget on larger consoles then porting was always gonna be viable. DLSS and FSR won't change anything. on consoles, FSR2.0 isn't a big leap because temporal upscaling has been a thing in the console space already

 

[Look over there]

Latest DF video on FSR 2.0 is very good, and FSR 2.0 looks great. Worth checking out for this thread because it (incidentally) makes a case for why DLSS is such a good win for Nintendo. To sum up

• DLSS performance is generally better, and likely to always be because tensor cores.
• FSR runs on any card - but that doesn't matter in the console space
• In comparison, current DLSS seems consistently better when working in performance mode.

And I'm getting the impression that at lower output resolutions, the gap between DLSS and FSR 2.0 gets wider?
I'm sitting here thinking, there are timelines here where, in terms of pure looks, Drake can meet or beat the Series S at its targeted resolutions within...ehh, we'll say anywhere between a fifth to a third of the power needed.
I'm also starting to lean more in the direction of 'the Series S is not gonna age well into the temporal-upscaling-becomes-required timeline, barring goalpost moving'.