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

[Dakhil]

I hope they don't label it as a pro nor successor - just call it the Super Switch and let people draw their own conclusions from the specs.

The problem is that Nintendo's probably going to be very vague when disclosing technical specs, similar with the Nintendo Switch. So, the average consumer can't really rely on Nintendo's disclosed technical specs to determine if the DLSS model* is a mid-gen refresh or a successor.

 

[ILikeFeet]

has flashbacks to the Wii U confusion with the general consumer

"Super Nintendo" is still widely in the public consciousness. It's not so much a crazy idea to call it a "Super Nintendo Switch"

 

[Skittzo]

has flashbacks to the Wii U confusion with the general consumer

When the Switch tablet is its own standalone console unit I don't think it could wind up causing that kind of confusion like the Wii U tablet did.

Nobody will be confused about a new Super Switch like they were the Wii U tablet controller.

 

[Jafin]

Oof, it was a joke, just a joke. "Super" is the one I've been wanting ever since we first heard concrete rumblings of a pro last year or the year before. Either Super Nintendo Switch or Nintendo Switch Super. (I'm rather partial to the latter)

 

[Hermii]

If it was my decision, I would go with “switch 2” plain and simple.


If it is indeed just a switch with better internals.

 

[Deleted member 645]

It’s getting announced tomorrow right…

 

[ACE+]

It’s getting announced tomorrow right…

Yes, alongside BotW2 and both coming out this summer.

 

[Alovon11]

Yes, alongside BotW2 and both coming out this summer.

Honestly wouldn't be surprised if Nintendo at least mentions "A bigger announcement coming this year" or something.

Dane likely won't come out until after Summer 2022, but they can tease/make a statement that it exists without giving many details until a proper reveal.

But that's the best case.

 

[Hosermess]

Stating it exists is the most important detail they could give. They won't do it tomorrow.

 

[Sol]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

 

[Hermii]

Honestly wouldn't be surprised if Nintendo at least mentions "A bigger announcement coming this year" or something.

Dane likely won't come out until after Summer 2022, but they can tease/make a statement that it exists without giving many details until a proper reveal.

But that's the best case.

“Hey everyone! If you’re on the fence on buying an oled, don’t!”

 

[Dakhil]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

Part of the problem is the engine Game Freak used for Pokémon Legends: Arceus based on what Digital Foundry said. And I don't think more powerful hardware is going to automatically fix all of the engine problems present in Pokémon Legends: Arceus.

Also, designing a no compromise hybrid console doesn't come cheap, especially given the inherent limitations of a hybrid console form factor.

 

[Hermii]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

Sony and MS took a significant loss on HW, and you know how Nintendo feels about that.

And they made compromises too. Like series x split memory, ssd capacity, and “only” doubling the ram from last gen. Compromises is the name of the game, when creating a mass market priced game console.

 

[ILikeFeet]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

DF even straight up says it's not a hardware issue. Pokemon is just a woeful lack of technical experience

 

[Alovon11]

“Hey everyone! If you’re on the fence on buying an oled, don’t!”

The people who care about a more powerful switch aren't buying the SWOLED already, especially if they are on the fence at this point.

And again, I said it's unlikely, a statement about "Something big in 2022" is likely the best we could get like I said.

 

[Dekuman]

Sony and MS took a significant loss on HW, and you know how Nintendo feels about that.

And they made compromises too. Like series x split memory, ssd capacity, and “only” doubling the ram from last gen. Compromises is the name of the game, when creating a mass market priced game console.

Hybrid console will always involve compromises. I think some people just carried away with all the speculation

I also think Nintendo will go for a break even pricing or a slight loss if they have to. They did it with the Switch, 3DS and Wii U.

 

[Instro]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

It seems like it's largely on GF, but I agree with the sentiment. Definitely been feeling pretty "done" with the current Switch hardware, and it impacts my excitement of upcoming announcements because they are probably not going to look or perform all that well.

 

[JoshuaJSlone]

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

It ain't like Series X comes with 64GB RAM, there are always limits.

 

[Sol]

It seems like it's largely on GF, but I agree with the sentiment. Definitely been feeling pretty "done" with the current Switch hardware, and it impacts my excitement of upcoming announcements because they are probably not going to look or perform all that well.

That's pretty much where I am. The hardware just feels limiting at this point, like the PS4 and Xbox One did around 2018-2019.

And unfortunately I don't think it will be until 2024 until we see the next console, unless I missed some new information in the last 40 pages or so.

 

[Tentacle-tropes]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

Having a more powerful Switch won’t necessarily mean that the game will run more stable. And, this device will have to cut corners somewhere to fit into their perceived price range and thermodynamics.

Edit: speaking of price it will probably be profitable day one or barely profitable. I don’t foresee Nintendo making a device they sell at a loss & needing games to cover for profitability.
WiiU: was probably profitable if you bought the standard version. After it was discontinued, leaving the premium version, it was really only profitable with a game purchase or two.
3DS: probably was barely profitable at its original price point. However, after the price cut was one of the reasons why they went into the red for the first time.
Switch: was profitable from basically say one and hasn’t really stopped. The only model that was barely profitable was the Lite and was what lead to some stock drama over not making more money when it came out.

 

[Alovon11]

Switch: was profitable from basically say one and hasn’t really stopped. The only model that was barely profitable was the Lite and was what lead to some stock drama over not making more money when it came out.

Yeah, I do say Dane likely will be more like the 3DS at launch, but able to handle the stress of the price point due to the OG Switch still being supported

Nintendo's most likely play is to have the gap between Dane's release and Eristra/Mariko EoL be high enough in order to have the costs for Dane to drop to the point a more entry-level model can make a profit for when OG Switches hit EoL.

 

[SiG]

I'm watching Digital Foundry's tech review of Pokemon Legends: Arceus, and all I could think about while watching was how much we need a more powerful Switch.

The game runs at a stable 720p and 30 DPS in portable mode, but can't even handle 1080p 30 DPS while docked, and dips down to 900p at 28-29 FPS at times.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

I honestly don't think it was an issue of specs quite frankly.

If anything Digital Foundry was saying even among Nintendo 1st party open-world titles on the same system, this looks like a big step down in presentation.

That's not even mentioning Kirby and the Forgotten Land, which seems to be semi-open...but even its world map area looks better presented than Arceus' environements.

To me, it strikes as having a game being very poorly optimized, or rather feels as if the team that worked on it was rather small.

 

[Deleted member 1324]

I think at this point we should expect a Summer 2023 reveal and September 2023 release due to o combination of the chip issues and how well the current Switch and newly released OLED is selling. Two years between models looks to be a good business strategy too.

 

[Vash_the_Stampede]

Hey all, new to this forum but I used to post on ResetEra (mostly the Nintendo tech thread and sales thread). It’s good to see many familiar faces/avatars.

I’m not a tech guy, I’m an investment analysts and Nintendo falls into my coverage.

I’ve been trying to figure out what Nintendo wants to do, and not what consumers want or techies dream of.

I think there are two scenarios here.

1. We get no new, higher performant hardware and Nintendo rides out the Mariko chip until it’s ready to launch the next gen. I don’t think we are getting next gen until holiday 2023 (the earliest). I also think everyone who believes in an iterative model (I.e., like the iPhone is wrong).

2. I have a new Nintendo Switch hardware theory. It’s sorta evil which makes me want to believe it’s real.

A 2022 Switch Pro based on Dane/Ampere but it will be just barely worth it. About 4x performance w/DLSS - enough to take games like BotW to 1440p docked and existing 1080p games to 4k. It will support more next gen/more demanding games but only at like 540p-720p docked.

I think they can sell 20m-40m units over 2- 3-years and then launch next gen in 2024-2025 on Nvidia 5nm.

The pros of this include:
-$400 price point; Gets consumers used to more expensive Nintendo hardware.
- Leaves plenty of room to market improved performance between Pro and Switch 2.
-Helps them resell more evergreen software and recoup any work they need to port/transfer games to the new hardware.

I wanted to add an addendum to this post.

In scenario #1 I said no new hardware until holiday 2023 at the earliest - but now I think that’s the latest it will occur.

I spend another day looking at my Nintendo model and I don’t think they can sell enough software to offset the decline in hardware that they are likely to see in this upcoming fiscal year and the one after that. They need a new piece of hardware.

Waiting until holiday 2023 would mean the console is ~7-years past launch and supporting it another 2-3 would mean a ~10-year life. That’s probably as good as it gets for hardware. The PS3 didn’t hit 10-years and PS4 will limp into year 10.

I’m still a little skeptical Dane will come out holiday 2022 as hardware should remain in the 20m-ish range and they should be able to match this fiscal years ~230m software units.

 

[brainchild]

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

 

[Dekuman]

I think at this point we should expect a Summer 2023 reveal and September 2023 release due to o combination of the chip issues and how well the current Switch and newly released OLED is selling. Two years between models looks to be a good business strategy too.

For what its worth the current nvidia roadmap is to move to Lovelace GTX 4000 series cards in the 2nd half of 2022, this opens up nvidia's current 8nm fabrication for Nintendo plus whatever extra capacity they may need, assuming rumors are true and Dane is on 8nm. As noted, the X1 was fabricated on an intermediate node between nvidia's GPU generations (28nm > 16nm) from the 700 series to 900 series, with the X1 fab at 20nm.

I don't think that's going to happen this time since we've heard nothing about it, and the timing for Lovelace lines up with a 2023 or late 2022 release for a Dane SoC powered Switch

 

[Alovon11]

I think at this point we should expect a Summer 2023 reveal and September 2023 release due to o combination of the chip issues and how well the current Switch and newly released OLED is selling. Two years between models looks to be a good business strategy too.

That is not how production works at all.

• Devkits have been out since 2020.
• Exclusives are timed for Fall 2022 or Early 2023.
• Orin is out soon/now (at least being put into cars)
• Switch Sales are already dipping trendwise.

You don't just tell devs "Oh yeah target this window" then less than a year from that window for these big titles say "Oh nevermind push it back another freaking year"

That's how you kill the third-party trust Nintendo has built up with the Switch.

 

[ReddDreadtheLead]

Concerning strictly the CPU in a vacuum?
Consider this: 'Dual channel', or 128-bit, 3200 MT/s DDR4 should be 51.2 GB/S, IIRC. In desktop space at least, I want to say that's actually already somewhat past the point of diminishing returns for current, 'general' computing for consumers. And I'm going to guess that a mobile CPU would probably need less bandwidth to keep fed compared to say, a desktop with more threads.

Of course, it's not in a vacuum, the RAM's shared with the GPU. So more bandwidth will help to potentially increase the amount leftover for the CPU
Although now I'm curious about whether there's a way to monitor how memory bandwidth usage is divided up in an SoC.

Yes as in strictly CPU in a vacuum (but also in a soc setting , )I was curious as the bandwidth situation could be enough to set this mobile chip from the other mobile chips.

From what I found, the A76 has a major bandwidth improvement of 90%

Memory bandwidth increased 90% relative to the A75.[6][7]According to ARM, the A76 is expected to offer twice the performance of an A73 and is targeted beyond mobile workloads. The performance is targeted at "laptop class", including Windows 10 devices,[8] competitive with Intel's Kaby Lake.[9]

ARM Cortex-A76 - Wikipedia

en.m.wikipedia.org

And here for the A77:

The ARM Cortex-A77 is a microarchitectureimplementing the ARMv8.2-A 64-bit instruction setdesigned by ARM Holdings' Austin design centre.[1]ARM announced an increase of 23% and 35% in integer and floating point performance, respectively. Memory bandwidth increased 15% relative to the A76.[1]

Wikiwand - ARM Cortex-A77

The ARM Cortex-A77 is a central processing unit implementing the ARMv8.2-A 64-bit instruction set designed by ARM Holdings' Austin design centre. ARM announced an increase of 23% and 35% in integer and floating point performance, respectively. Memory bandwidth increased 15% relative to the A76.

www.wikiwand.com

While the A78C has this according to Techspot:

The A78C can be paired with up to 8 MB of L3 cache, and depending on manufacturers’ preferences, each core can have either 256 or 512 kB of L2 cache. The L1 instruction and data caches can both be either 32 or 64 kB. The processor’s maximum memory bandwidth is 60 GB/s.

Arm's Cortex-A78C is a powerful new processor core for laptops

About four years ago, Microsoft announced that they were partnering with Qualcomm to bring the full Windows 10 experience to Arm-based mobile processors. It's been a bit...

www.techspot.com

So it seems like the A78C (and A78 by extension) are >2.18x the memory bandwidth of what could be in the Switch with the 25GB/s situation. Mobile chips don’t really go to 128-bit or close to 90-100GB/s, so this switch could have a leg up on the other mobile devices with memory bandwidth, but I’m not sure by how much more performant it would be.

And a lot of the Geekbench5 testings done that get posted typically are with 64-bit 34-51GB/s(?) memory. Granted, they don’t want to really fetch from the RAM as often due to it requiring more energy to do so.

But this is a console so it has to contend with different situations than the others.

I’m not sure how it would work though since it is system shared, and not separate and only for the CPU or only for the GPU.

 

[minimalism]

That is not how production works at all.

• Devkits have been out since 2020.
• Exclusives are timed for Fall 2022 or Early 2023.
• Orin is out soon/now (at least being put into cars)
• Switch Sales are already dipping trendwise.

You don't just tell devs "Oh yeah target this window" then less than a year from that window for these big titles say "Oh nevermind push it back another freaking year"

That's how you kill the third-party trust Nintendo has built up with the Switch.

Peeps tend not to understand how manufacturing works. I work in long-term products (>30 year lifetimes) and the detail design phase of products happens nearly a decade before production ever starts. As a result, there's a huge gap between current tech 10 years ago versus now. Yet, all the requirements and procurement details are still written to decade old requirements with no updates because that would cause scope creep (aka: $$$).

That being said, consoles are no where near my production time frame but still need a line-in-the-sand for a requirement agreement on specs. You can't ever expect scope to keep creeping year by year chasing progress and still expect corporate to pay for it. If there are devkits out in the wild now the specs are set in stone and aren't changing; chip shortage be damned.

 

[ShadowFox08]

Steam Deck battery can drain in under 2 hours in some cases | Digital Trends

Depending on what game is being played and the settings it's running at, the Steam Deck's battery life can vary drastically.

www.digitaltrends.com

So tests for battery life were taken for Steam Deck and one person noticed the battery life went as little as 1.5-2 hrs for Devil May Cry 5 at 50% brightness and audio cut off. High settings with vsync made it last 2 hrs. Increasing the settings (doesn't say specifically what) and turning vsync off made it last 87 minutes.

Interesting..

 

[Alovon11]

Peeps tend not to understand how manufacturing works. I work in long-term products (>30 year lifetimes) and the detail design phase of products happens nearly a decade before production ever starts. As a result, there's a huge gap between current tech 10 years ago versus now. Yet, all the requirements and procurement details are still written to decade old requirements with no updates because that would cause scope creep (aka: $$$).

That being said, consoles are no where near my production time frame but still need a line-in-the-sand for a requirement agreement on specs. You can't ever expect scope to keep creeping year by year chasing progress and still expect corporate to pay for it. If there are devkits out in the wild now the specs are set in stone and aren't changing; chip shortage be damned.

Yeah, I will say for consoles there is a little bit of wiggle room for partners to have influence but that extends to storage/RAM usually, not the SoC.

So the timing and specs for Dane dev-wise have been reigned in for a bit by now and pushing things back a year would do more damage than it's worth

 

[Vash_the_Stampede]

For what its worth the current nvidia roadmap is to move to Lovelace GTX 4000 series cards in the 2nd half of 2022, this opens up nvidia's current 8nm fabrication for Nintendo plus whatever extra capacity they may need, assuming rumors are true and Dane is on 8nm. As noted, the X1 was fabricated on an intermediate node between nvidia's GPU generations (28nm > 16nm) from the 700 series to 900 series, with the X1 fab at 20nm.

I don't think that's going to happen this time since we've heard nothing about it, and the timing for Lovelace lines up with a 2023 or late 2022 release for a Dane SoC powered Switch

I’m coming around to this view, slowly.

 

[minimalism]

Yeah, I will say for consoles there is a little bit of wiggle room for partners to have influence but that extends to storage/RAM usually, not the SoC.

So the timing and specs for Dane dev-wise have been reigned in for a bit by now and pushing things back a year would do more damage than it's worth

Agree, I was oversimplifying some stuff. There are certainly things that can be changed down the road for 'free' (aka at the cost of an Engineering Change Order), but the guts of it are absolutely not changing come hell or high water. People who haven't worked in manufacturing don't get how process driven it is and even small changes cost tons of money.

For example, at my company, a revision change costs in excess of 100k by the time all the red tape is cut through. That's without having to go to the vendor/customer and have them process the change and their associated costs. But that's either here nor there, just proving why once you have your requirements set you don't deviate.

 

[Alovon11]

Agree, I was oversimplifying some stuff. There are certainly things that can be changed down the road for 'free' (aka at the cost of an Engineering Change Order), but the guts of it are absolutely not changing come hell or high water. People who haven't worked in manufacturing don't get how process driven it is and even small changes cost tons of money.

For example, at my company, a revision change costs in excess of 100k by the time all the red tape is cut through. That's without having to go to the vendor/customer and have them process the change and their associated costs. But that's either here nor there, just proving why once you have your requirements set you don't deviate.

Yeah, I am saying the same thing applies for Dane more or less.
The vendor point being more apt as unlike the Steam Deck which is solely sold and vendored by Valve themselves, the console Dane is in will have to contend with marketing and supply contracts from Dozens of Retailers across the world.

Not to mention the advertising firms that Nintendo would hire to make their ads to advertise the console in some regions.

There are way more moving parts in the launch for Dane that make the idea of pushing it a year back from 2H 2022/1H 2023 to 2H 2023/1H 2024 "because shortages" impossible realistically in my eyes.

 

[Deleted member 645]

Luck isn’t needed just not going to happen

 

[minimalism]

Yeah, I am saying the same thing applies for Dane more or less.
The vendor point being more apt as unlike the Steam Deck which is solely sold and vendored by Valve themselves, the console Dane is in will have to contend with marketing and supply contracts from Dozens of Retailers across the world.

Not to mention the advertising firms that Nintendo would hire to make their ads to advertise the console in some regions.

There are way more moving parts in the launch for Dane that make the idea of pushing it a year back from 2H 2022/1H 2023 to 2H 2023/1H 2024 "because shortages" impossible realistically in my eyes.

I can see stuff like that pushing out just due to engineering missing milestone dates. Something that someone wrote a schedule for 2 years ago is being blown due to turnover/inexperience/inaccurate assumptions too. I've been a design engineer in that position (and am in a minus slack project now, unfortunately) and it's just so gosh darn easy to be confident on issues and designs you haven't touched outside conceptually.

Personally I think we're looking at the test units not meeting test specs and Ninty fighting a chip shortage + engineering issues trying to make it hit said target specs.

Your first hint at that was the OLED Switch. Clearly the next-gen Switch planned to use all those parts plus a new chip so they started ordering those new, low risk, easy to procure, low failure, parts en masse. Then the snag hit and they scrambled and Frankensteined their already procured (long term-contract in place) parts into a new SWOLED unit so they wouldn't eat the cost of the unused components.

I can certainly be wrong but my experience as a design engineer just connects too many dots (and experiences) in this journey that's it's just eeriely close to what's I've been through.

 

[SiG]

I feel that the imminent release of the Steam Deck just makes people's expectation of a Switch successor to be much higher, and they're also expecting it to be priced competitively. Keep in mind that Valve is took pains (i.e. sold at a loss) to get their console within those price brackets...

So yeah, I'm expecting new people here to be swarming with posts about "Is the next Switch going to be as powerful as the Steam Deck" to which I'll say right away "Probably not". To which said newcomers will probably be "Bah! Switch sucks, I hope they become 3rd party etc." and the cycle continues.

At the same time, I'm glad Valve are releasing their own affordable gaming PCs (the Deck is pretty much considered a PC) that won't be scalped for GPU mining. Then again, miners/scalpers will probably find a way or something...

 

[Alovon11]

I feel that the imminent release of the Steam Deck just makes people's expectation of a Switch successor to be much higher, and they're also expecting it to be priced competitively. Keep in mind that Valve is took pains (i.e. sold at a loss) to get their console within those price brackets...

So yeah, I'm expecting new people here to be swarming with posts about "Is the next Switch going to be as powerful as the Steam Deck" to which I'll say right away "Probably not". To which said newcomers will probably be "Bah! Switch sucks, I hope they become 3rd party etc." and the cycle continues.

At the same time, I'm glad Valve are releasing their own affordable gaming PCs (the Deck is pretty much considered a PC) that won't be scalped for GPU mining. Then again, miners/scalpers will probably find a way or something...

Well, the thing is when portable, yeah SD will likely win.

But docked Dane will likely be around the 1050Ti and that's before DLSS or NVIDIA's uArch efficiency gains over Pascal are put in place.

That singlehandedly beats the Steam Deck.

 

[Look over there]

Yes as in strictly CPU in a vacuum (but also in a soc setting , )I was curious as the bandwidth situation could be enough to set this mobile chip from the other mobile chips.

From what I found, the A76 has a me doy bandwidth improvement of 90%

ARM Cortex-A76 - Wikipedia

en.m.wikipedia.org

And here for the A77:

Wikiwand - ARM Cortex-A77

The ARM Cortex-A77 is a central processing unit implementing the ARMv8.2-A 64-bit instruction set designed by ARM Holdings' Austin design centre. ARM announced an increase of 23% and 35% in integer and floating point performance, respectively. Memory bandwidth increased 15% relative to the A76.

www.wikiwand.com

While the A78C has this according to Techspot:

Arm's Cortex-A78C is a powerful new processor core for laptops

About four years ago, Microsoft announced that they were partnering with Qualcomm to bring the full Windows 10 experience to Arm-based mobile processors. It's been a bit...

www.techspot.com

So it seems like the A78C (and A78 by extension) are >2.18x the memory bandwidth of what could be in the Switch with the 25GB/s situation. Mobile chips don’t really go to 128-bit or close to 90-100GB/s, so this switch could have a leg up on the other mobile devices with memory bandwidth, but I’m not sure by how much more performant it would be.

And a lot of the Geekbench5 testings done that get posted typically are with 64-bit 34-51GB/s(?) memory. Granted, they don’t want to really fetch from the RAM as often due to it requiring more energy to do so.

But this is a console so it has to contend with different situations than the others.

I’m not sure how it would work though since it is system shared, and not separate and only for the CPU or only for the GPU.

It's hard to say since we don't have base numbers to work off of. So we're unsure if 34-51 GB/s is sufficient or not to feed the A78 cores to their max potential already.
And with the GPU in the mix, we don't know how the 25 GB/s in the OG Switch is being split now, nor how it'd be split in Dane. Although it's probably likely that CPU still gets more to work with compared to the OG model.

That said, I don't expect ARM cores to have all that high bandwidth needs. I mean, 3900x/Zen 2's memory controller numbers are all at ~25 GB/s or less while 9900k/Skylake's not all that far off. And from here, 1065G7/Ice Lake's memory controller numbers are 30 GB/s and under, while 1185G7/Tiger Lake's ~35 GB/s and under.

 

[RainTech]

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

The hybrid concept alone puts alot of restrictions on how much they can do.... Wouldn't get my hopes up.

 

[Crusters]

Well, the thing is when portable, yeah SD will likely win.

But docked Dane will likely be around the 1050Ti and that's before DLSS or NVIDIA's uArch efficiency gains over Pascal are put in place.

That singlehandedly beats the Steam Deck.

I've heard about this. But aside from being an orin family product, I haven't really gotten my hands on much that's solid. I will be going through this thread to catch up, so sorry if I am really out of the loop....

Around a 1050ti hybrid is impressive to me. But how.... Palpable is that? Do we actually have an idea of the render config on this Dane?

 

[ShadowFox08]

I think at this point we should expect a Summer 2023 reveal and September 2023 release due to o combination of the chip issues and how well the current Switch and newly released OLED is selling. Two years between models looks to be a good business strategy too.

I can't wait for a 8nm node in 2023!!

I feel that the imminent release of the Steam Deck just makes people's expectation of a Switch successor to be much higher, and they're also expecting it to be priced competitively. Keep in mind that Valve is took pains (i.e. sold at a loss) to get their console within those price brackets...

So yeah, I'm expecting new people here to be swarming with posts about "Is the next Switch going to be as powerful as the Steam Deck" to which I'll say right away "Probably not". To which said newcomers will probably be "Bah! Switch sucks, I hope they become 3rd party etc." and the cycle continues.

At the same time, I'm glad Valve are releasing their own affordable gaming PCs (the Deck is pretty much considered a PC) that won't be scalped for GPU mining. Then again, miners/scalpers will probably find a way or something...

I don't think Nintendo is worried, but the more competition the better for sure.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

If Nintendo can be as performant as they can while maintaining the same battery life (2.5-4hrs for demanding games) as V1 switch, I'll be more than happy. I think matching Steam Deck's GPU of 1.5-1.6 in on paper specs will be as high as they can get in TFLOPs (2.6x as less as Orion Nx's 4 TFLOPs), but I think it will be doable.

Somewhat unrelated. Maybe it's just me, I didn't know Orion NX was confirmed with the Jetson name on front and 10, 15 and 25 watt modes with 12GB and 102GB/ bandwidth?
I thought before the big event last year we had that info on it with those 3 power watt modes, but it disappeared and it was some thing else (I remember 8 core A78s and half the GPU power). I dunno. Or more info just got released recently (module?)... At least the 8 core CPU has the same 2MB L2 + 4MB L3.

NVIDIA Jetson Orin

Next-level AI performance for next-gen robotics.

developer.nvidia.com

NVIDIA Jetson AGX Orin

Next-level AI performance for next-gen robotics.

www.nvidia.com

NVIDIA Jetson Orin

Next-level AI performance for next-gen robotics.

www.nvidia.com

70mm x 45mm size for the module vs 100mmx 86mm for AGX Orion also.

 

[Alovon11]

I've heard about this. But aside from being an orin family product, I haven't really gotten my hands on much that's solid. I will be going through this thread to catch up, so sorry if I am really out of the loop....

Around a 1050ti hybrid is impressive to me. But how.... Palpable is that? Do we actually have an idea of the render config on this Dane?

8SMs is the most likely config GPU wise when looking at Orin NX which is the binned version of Orin AGX.

That would give it 1024 CUDA cores, 4 RT Cores (Of unknown generation), and 32 3rd-Gen Tensor Cores.

Orin also increased the L1 Cache by 50% and the L2 Cache by double over Ampere so Dane would have that much more cache versus an 8SM Ampere part.

Looking at the lowest end Ampere Card, the NVIDIA A2, we can tell that a 10SM Ampere part with unoptimized drivers at 1.7Ghz boost clocks can Surpass the 1050Ti by a fair bit, getting quite close to the PS4 Pro's GPU candidates (The HD 7970 Ghz for a low-end estimate, the RX 470 for a higher-end estimate for that console)

Considering Ampere's critical flaw with it's FLOP efficiency is the lack of memory bandwidth to feed the doubling of CUDA cores over Turing, we could expect that relatively massive increase in Cache Orin provides to push Ampere closer back to Turing FLOP Efficiency (For reference, a Turing TFLOP is nearly worth 1.5 Ampere TLFOPs)

So the 4.5 TFLOP A2 at 1.7ghz peak would be a 3.1TFLOP Turing GPU.

And we also have to consider that Ampere has double the core count over Turing so if that cache can keep every core absolutely fed then Orin could push that equitable number even lower.

So it is perfectly reasonable to assume in my opinion that 8SMs of Orin would at least match 10SMs of Ampere.
Which then it comes down to clock speed which while Orin caps at 1Ghz, Dane doesn't necessarily need to.

But even if it caps at 1Ghz, we know how Ampere performs at 1Ghz for reference, with the Laptop RTX 3050 having a boost clock of 1057Mghz at 16SMs.

So considering the 8SM Orin = 10SM Ampere at the same clocks guess, we can use that to figure out that 16SM Ampere would match to 13SMs of Orin at the same clocks at the worst-case scenario for that uplift.


The RTX 3050 Laptop is still ahead of the 1050Ti Desktop at this point, so I say that dropping down to 8SMs for Dane considering how Ampere seems to scale on SM count worse than previous uArchs, The lower-bound for Dane at 1Ghz would likely be a GTX 960.

Which the 960 is no slouch either.

And ofc, the absolute worst case for Docked Dane would be GTX 1050 level performance which would still pit it ahead of the Steam Deck when docked because better RT and DLSS.

All configs here put it well ahead of the OG PS4 GPU btw

 

[Crusters]

8SMs is the most likely config GPU wise when looking at Orin NX which is the binned version of Orin AGX.

That would give it 1024 CUDA cores, 4 RT Cores (Of unknown generation), and 32 3rd-Gen Tensor Cores.

Orin also increased the L1 Cache by 50% and the L2 Cache by double over Ampere so Dane would have that much more cache versus an 8SM Ampere part.

Looking at the lowest end Ampere Card, the NVIDIA A2, we can tell that a 10SM Ampere part with unoptimized drivers at 1.7Ghz boost clocks can Surpass the 1050Ti by a fair bit, getting quite close to the PS4 Pro's GPU candidates (The HD 7970 Ghz for a low-end estimate, the RX 470 for a higher-end estimate for that console)

Considering Ampere's critical flaw with it's FLOP efficiency is the lack of memory bandwidth to feed the doubling of CUDA cores over Turing, we could expect that relatively massive increase in Cache Orin provides to push Ampere closer back to Turing FLOP Efficiency (For reference, a Turing TFLOP is nearly worth 1.5 Ampere TLFOPs)

So the 4.5 TFLOP A2 at 1.7ghz peak would be a 3.1TFLOP Turing GPU.

And we also have to consider that Ampere has double the core count over Turing so if that cache can keep every core absolutely fed then Orin could push that equitable number even lower.

So it is perfectly reasonable to assume in my opinion that 8SMs of Orin would at least match 10SMs of Ampere.
Which then it comes down to clock speed which while Orin caps at 1Ghz, Dane doesn't necessarily need to.

But even if it caps at 1Ghz, we know how Ampere performs at 1Ghz for reference, with the Laptop RTX 3050 having a boost clock of 1057Mghz at 16SMs.

So considering the 8SM Orin = 10SM Ampere at the same clocks guess, we can use that to figure out that 16SM Ampere would match to 13SMs of Orin at the same clocks at the worst-case scenario for that uplift.


The RTX 3050 Laptop is still ahead of the 1050Ti Desktop at this point, so I say that dropping down to 8SMs for Dane considering how Ampere seems to scale on SM count worse than previous uArchs, The lower-bound for Dane at 1Ghz would likely be a GTX 960.

Which the 960 is no slouch either.

And ofc, the absolute worst case for Docked Dane would be GTX 1050 level performance which would still pit it ahead of the Steam Deck when docked because better RT and DLSS.

All configs here put it well ahead of the OG PS4 GPU btw

8 SM'S? 1024 Cuda Cores? So you think it's going for exactly half a GCU?

I'd be pretty pleased with 6/around Durango Xbox shader counts honestly.

Any chance anyone have an idea of the texture mappers and rops that would be paired with this config off the top of their heads?

 

[Hermii]

I can see stuff like that pushing out just due to engineering missing milestone dates. Something that someone wrote a schedule for 2 years ago is being blown due to turnover/inexperience/inaccurate assumptions too. I've been a design engineer in that position (and am in a minus slack project now, unfortunately) and it's just so gosh darn easy to be confident on issues and designs you haven't touched outside conceptually.

Personally I think we're looking at the test units not meeting test specs and Ninty fighting a chip shortage + engineering issues trying to make it hit said target specs.

Your first hint at that was the OLED Switch. Clearly the next-gen Switch planned to use all those parts plus a new chip so they started ordering those new, low risk, easy to procure, low failure, parts en masse. Then the snag hit and they scrambled and Frankensteined their already procured (long term-contract in place) parts into a new SWOLED unit so they wouldn't eat the cost of the unused components.

I can certainly be wrong but my experience as a design engineer just connects too many dots (and experiences) in this journey that's it's just eeriely close to what's I've been through.

I find your posts very interesting. What kind of products are you working on if you dont mind me asking? The only thing I can think of which has 30 years lifetime and 10 years design is military,

 

[Celine]

Steam Deck battery can drain in under 2 hours in some cases | Digital Trends

Depending on what game is being played and the settings it's running at, the Steam Deck's battery life can vary drastically.

www.digitaltrends.com

So tests for battery life were taken for Steam Deck and one person noticed the battery life went as little as 1.5-2 hrs for Devil May Cry 5 at 50% brightness and audio cut off. High settings with vsync made it last 2 hrs. Increasing the settings (doesn't say specifically what) and turning vsync off made it last 87 minutes.

Interesting..

Less than 1.5 hrs? that's not great.
IMO the bare minimum a battery life should last is the duration of a whole gaming session (the reasonably longest you'll end up playing in one time), you can likely recharge the unit after.
A gaming session longer than 1.5 hrs isn't unreasonable for me.

 

[ShadowFox08]

I feel that the imminent release of the Steam Deck just makes people's expectation of a Switch successor to be much higher, and they're also expecting it to be priced competitively. Keep in mind that Valve is took pains (i.e. sold at a loss) to get their console within those price brackets...

So yeah, I'm expecting new people here to be swarming with posts about "Is the next Switch going to be as powerful as the Steam Deck" to which I'll say right away "Probably not". To which said newcomers will probably be "Bah! Switch sucks, I hope they become 3rd party etc." and the cycle continues.

At the same time, I'm glad Valve are releasing their own affordable gaming PCs (the Deck is pretty much considered a PC) that won't be scalped for GPU mining. Then again, miners/scalpers will probably find a way or something...

I hope Nintendo uses this as motivation to be more competitive.

I really hope Nintendo does what Microsoft and Sony did with their new consoles, and goes all out in truly creating the best and most technically advanced machine they can make. No cut corners.

If Nintendo can be as performant as they can while maintaining the same battery life as V1 switch, I'll be more than happy. I think matching Steam Deck's GPU of 1.5-1.6 in on paper specs will be as high as they can get in TFLOPs (2.6x as less as Orion Nx's 4 TFLOPs), but I think it will be doable. And they are cutting RAM from Orion NX from 32GB(?) to 12, as well has a 6-8 core CPU--likely performing at around 1-1.5GH per core. That should get it into a a 15 watt range I think.

Less than 1.5 hrs? that's not great.
IMO the bare minimum a battery life should last is the duration of a whole gaming session (the reasonably longest you'll end up playing in one time), you can likely recharge the unit after.
A gaming session longer than 1.5 hrs isn't unreasonable for me.

Might as well call it 1.5 hrs. heh

That's true 1.5hrs is a bit low. I get that it's customized like a PC with the highest settings (which we don't seem to know about in detail), but I wonder what the CPU and GPU clocks were at that moment... Devil May Cry 5 is s PS4 level game

 

[Pokemaniac]

I can see stuff like that pushing out just due to engineering missing milestone dates. Something that someone wrote a schedule for 2 years ago is being blown due to turnover/inexperience/inaccurate assumptions too. I've been a design engineer in that position (and am in a minus slack project now, unfortunately) and it's just so gosh darn easy to be confident on issues and designs you haven't touched outside conceptually.

Personally I think we're looking at the test units not meeting test specs and Ninty fighting a chip shortage + engineering issues trying to make it hit said target specs.

Your first hint at that was the OLED Switch. Clearly the next-gen Switch planned to use all those parts plus a new chip so they started ordering those new, low risk, easy to procure, low failure, parts en masse. Then the snag hit and they scrambled and Frankensteined their already procured (long term-contract in place) parts into a new SWOLED unit so they wouldn't eat the cost of the unused components.

I can certainly be wrong but my experience as a design engineer just connects too many dots (and experiences) in this journey that's it's just eeriely close to what's I've been through.

The theory that Switch OLED was a last minute sawp is definitely something that's been floating around, but there's not really any evidence to back it up. The hardware config indicated it would use the Mariko SoC since the moment it appeared in public firmware builds in early 2020.

 

[SiG]

I hope Nintendo uses this as motivation to be more competitive.

I'm not quite sure about that. The one thing Nintendo seems to actively avoid is the spec war, and are more about bringing something unique in terms of the gaming experience. i.e. non-standardized gaming experiences.

I do wonder if the current situation will make them strike a better deal somehow. Worst case is that they do stick to this highly plausible 8nm "Dane" chip, which could still pull its weight given its size class. The issue is going to be pricing: If they price too agressively, then the Steam Deck will obviously offer more incentive for users within the same price bracket (as it is essentially a PC), but at the same time they don't want to kill the Switch momentum.

They key component will be a unique gimmick, and a certain segment (GAF/ERA) most notably dislike these "gimmicks". I'm pretty sure not much attention will be given to that segment (much to their chagrin).

tl;dr - Nintendo will find other ways to make the new device competitive. Some people won't have it because they're too busy with the spec measuring contest.

 

[Hermii]

I'm not quite sure about that. The one thing Nintendo seems to actively avoid is the spec war, and are more about bringing something unique in terms of the gaming experience. i.e. non-standardized gaming experiences.

Yep. They will say it’s 4k capable, but be quiet about exactly how that is achieved.They will only give the most superficial spec info. Let the games do the talking.