Based on how tech has progressed since the Switch, how powerful could a Switch Pro in early 2021 actually be?

z0m3le · Apr 7, 2020

I think it's worth talking about transistor density via the different processes since Switch came out, so people can understand a bit better what Nintendo and Nvidia have available to them for performance in a ~100mm² SoC.

First we will talk about Tegra X1 on 20nm, Tegra X1 is 121mm² with 2 Billion transistors which is about 16.5 Million transistors per mm², mobile parts like these sometimes leave room between the transistor for less heat at higher clocks, the actual density of 20nm is a little over 20 Million transistors/mm². It is worth noting that the density might not be as high to avoid the power leakage flaw found in 20nm, this shouldn't be an issue with newer process nodes as 3D transistors took over after this and solved that issue. I'll go with a pessimistic look at density to give us an idea of how big a difference this performance jump could be.

2015 TSMC 16nm: 28.2M/mm²
2.8 Billion transistors @100mm² SoC

2015 Samsung 14nm: 32.8M/mm²
3.3 Billion transistors @100mm² SoC

2017 TSMC 10nm: 48.1M/mm²
4.8 Billion transistors @100mm² SoC

2017 Samsung 10nm: 51.8M/mm²
5.2 Billion transistors @100mm² SoC

2018 Samsung 8nm: 61.2M/mm²
6.1 Billion transistors @100mm² SoC (Scorpio has a 7 Billion transistor SoC for comparison)

2018 TSMC 7nm: 96.5M/mm²
9.6 Billion transistors @100mm² SoC

2019 Samsung 7nm: 95.3M/mm²
9.5 Billion transistors @100mm² SoC

2020? Samsung 5nm: 127M/mm²
12.7 Billion transistors @100mm² SoC

2020 TSMC 5nm: 173M/mm² (in next iPhone)
17.3 Billion transistors @100mm² SoC

2022 3nm should offer >200M/mm2

We are looking at >6x the density for a Switch 2 at 121mm² however I think they will go with a smaller SoC, between 80mm and 100mm, for something around 10 Billion transistors.

8 core A78 and 4 core A55 as well as 1280 Cuda core GPU with tensor and rt cores should fit within the transistor and size budget with room left over on samsung's inferior 5nm process, but I think Nvidia will go with the more dense TSMC 5nm for any Tegra chip, as they can spread the transistors apart farther for cooler high clocks.

Maple said:
Is anyone else worried that Nintendo will somehow screw things up with their next console? A quick look at their history reveals that they really don't like simply releasing a more powerful version of the same system and adding a number to the end of it. They really want to fundamentally change how people use the system with each and every generation. Gamecube --> Wii --> Wii U. GBA --> DS --> 3DS. They always come up with a new "gimmick" they feel the need push with each new system.

I would love for nothing more than for Nintendo to release a Switch 2 - same design, same concept, fully BC with existing Switch library, with an Ampere based GPU, 8 GB of RAM, powerful ARM cores, OLED display, Bluetooth Audio, etc. But it seems like this goes against what they've been doing as a company for so long. They want to reinvent the wheel with each new generation.

Also thought this was worth addressing. Nintendo has 1 very clear pattern over it's ~40 years in the game industry.

They innovate and then refine, in a 2 step cycle that then repeats:
NES -> SNES => N64 -> Gamecube => Wii -> Wii U => Switch
GB/C -> GBA => DS -> 3DS => Switch
They come up with something, refine it in a new product, then invent something else, it even happens in their games:
Super Mario Bros -> Lost Levels => Super Mario Bros 3 -> Super Mario World => Mario 64 -> Mario Sunshine => Mario Galaxy -> Mario Galaxy 2 => Super Mario 3D Land -> Super Mario 3D World => Super Mario Odyssey

Nintendo also circles back around to older ideas to help innovate, but there is no doubt in my mind that the successor to the Switch will be a refinement on the Switch, and yes it might have a new gimmick like Wii U or 3DS had, it will still see a linear upgrade to the current Switch. That is the one consistent pattern I've been able to see from Nintendo throughout the years.

MikeE21286 · Apr 8, 2020

I'm guessing a PS4 or XB1 level. I don't think a PS4 Pro level is likely.

I always go minus 8 years from the date and see what console we're doing then in order to equate what a handheld can do. Of course this is completely unscientific and probably oversimplifying and/or inaccurate

Deleted member 49166 · Apr 8, 2020

MikeE21286 said:
I always go minus 8 years from the date and see what console we're doing then in order to equate what a handheld can do. Of course this is completely unscientific and probably oversimplifying and/or inaccurate

indeed it is ;)

ILikeFeet · Apr 11, 2020

a new low-end Nvidia mobile GPU has been spotted. the article linked is NotebookCheck's exclusive info about a TU117-based MX gpu (nvidia's lowest end mobile line). the TU117 is where the GTX 1650 lies (as low as 896 cores).

NVIDIA planning a Turing-based MX350 successor to take on Intel Tiger Lake's Xe DG1 iGPU, will be a GTX 1650-class chip with PCIe Gen4 support

NVIDIA is working on a Turing-based successor to the upcoming MX350 dGPU, according to our sources familiar with the matter. The new MX GPU will be based on the TU117 GPU found in the GTX 1650 and will be the first Turing card to feature PCIe Gen4 support. The TU117 MX GPU is designed to take on...

www.notebookcheck.net

I bring this up because the Switch's gpu configuration has been in laptops as the MX110. now, I don't believe that a Switch 2 would be Turing, but I think the existence of this gpu helps set the minimum expectation

Governergrimm · Apr 11, 2020

nsilvias said:
its not about how powerful it could be, its about how long the battery would last with current battery tech. we have phones with graphical performance near that of consoles but we have no way of keeping them running for long periods of time at high power

It's also about cost. Phones with that level of performance aren't cheap. As great as Nintendo is. They are going to have a gen if they launch at powerful portable tech prices.

ILikeFeet · Apr 11, 2020

Governergrimm said:
It's also about cost. Phones with that level of performance aren't cheap. As great as Nintendo is. They are going to have a gen if they launch at powerful portable tech prices.

phones have more expensive everything else. for all the talk of phone gpus there are few games that even take advantage of them (many of them, I've seen, were Asian MMOs running UE4). but stuff like modems, screens and cameras (upwards to $100 for the camera alone)

Why are modern phones so expensive?

Why are we well on our way to hit the $1500 phone price at the top end? Big brand CEOs blame 5G and periscope zoom.

www.phonearena.com

SiG · Apr 11, 2020

MikeE21286 said:
I'm guessing a PS4 or XB1 level. I don't think a PS4 Pro level is likely.

I always go minus 8 years from the date and see what console we're doing then in order to equate what a handheld can do. Of course this is completely unscientific and probably oversimplifying and/or inaccurate

The "generations behind" sentiment for Nintendo hardware is completely overblown. Also, this seems to mostly apply for their partnership during the AMD years.

Governergrimm said:
It's also about cost. Phones with that level of performance aren't cheap. As great as Nintendo is. They are going to have a gen if they launch at powerful portable tech prices.

Depends on the tech they cut from a phone. Most phones have high end cameras and OLED screens that contribute significantly to their price.

Governergrimm · Apr 11, 2020

ILikeFeet said:
phones have more expensive everything else. for all the talk of phone gpus there are few games that even take advantage of them (many of them, I've seen, were Asian MMOs running UE4). but stuff like modems, screens and cameras (upwards to $100 for the camera alone)

Why are modern phones so expensive?

Why are we well on our way to hit the $1500 phone price at the top end? Big brand CEOs blame 5G and periscope zoom.

www.phonearena.com

Sure but cut the cost in half. It's still expensive for a console. The comparison isn't 1:1 obviously but high tech and portability isn't cheap. That's a huge expense. If we consider upcoming consoles at an estimated $500 a console. We MAY see PS4 Pro level visuals but that won't be cheap. That also means they would need a better display than what's on the switch and that would drive up price too. Portability comes at a cost, either through financial or performance.

ILikeFeet · Apr 11, 2020

Governergrimm said:
Sure but cut the cost in half. It's still expensive for a console. The comparison isn't 1:1 obviously but high tech and portability isn't cheap. That's a huge expense. If we consider upcoming consoles at an estimated $500 a console. We MAY see PS4 Pro level visuals but that won't be cheap. That also means they would need a better display than what's on the switch and that would drive up price too. Portability comes at a cost, either through financial or performance.

I think you're missing the point. the SoCs aren't why phones are so expensive. it's not a "just cut it in half" kinda thing. phones have higher BoMs than even next gen consoles will have. there's a lot going on in these things. matching the performance of a phone SoC isn't gonna be the hardest thing to accomplish at near Switch-levels of cost

Governergrimm · Apr 11, 2020

SiG said:
The "generations behind" sentiment for Nintendo hardware is completely overblown. Also, this seems to mostly apply for their partnership during the AMD years.

Depends on the tech they cut from a phone. Most phones have high end cameras and OLED screens that contribute significantly to their price.

Cameras sure but if they are chasing 1440p or more then the screens sizes will be just as expensive. If they push power they will push res and that will drive up screen cost and cooling.

ILikeFeet · Apr 11, 2020

Governergrimm said:
Cameras sure but if they are chasing 1440p or more then the screens sizes will be just as expensive. If they push power they will push res and that will drive up screen cost and cooling.

why would Nintendo choose anything higher than a 1080p screen though? pushing power means they can cram more effects into 1080p. less power overall, cheaper, and at such a small screen size, the differences aren't as perceptable

SiG · Apr 11, 2020

Governergrimm said:
Cameras sure but if they are chasing 1440p or more then the screens sizes will be just as expensive. If they push power they will push res and that will drive up screen cost and cooling.

I have a feeling they'll be sticking with 720p screens but if small 1080p IGZO screens end up being cheaper due to their ubiquity we could potentially see that instead. I don't think they'll be chasing current phone/tablet resolutions when all that rendering power can be utilized elsewhere.

Governergrimm · Apr 11, 2020

ILikeFeet said:
I think you're missing the point. the SoCs aren't why phones are so expensive. it's not a "just cut it in half" kinda thing. phones have higher BoMs than even next gen consoles will have. there's a lot going on in these things. matching the performance of a phone SoC isn't gonna be the hardest thing to accomplish at near Switch-levels of cost

No, I get that. All I'm saying is that portability requires trade offs. A 2021 switch isn't going to be putting out PS4 pro/Xbox 1X at any kind of reasonable price.

It depends on what kind specs we are nailing it down to as well. Are we saying 4k, 1440p, 60fps, 120 fps? If everyone is using a different baseline then we will all have different opinions on possible cost.

Governergrimm · Apr 11, 2020

SiG said:
I have a feeling they'll be sticking with 720p screens but if small 1080p IGZO screens end up being cheaper due to their ubiquity we could potentially see that instead. I don't think they'll be chasing current phone/tablet resolutions when all that rendering power can be utilized elsewhere.

That's exactly what I mean. If we are saying PS4/X1 levels then I assume 1080p levels. If we say PS4 pro/X1X then I assume 1440-4k. Those screens are not cheap and neither is the power they draw in portable form. Same thing with fps. I guess I am assuming when we say PS4 levels that we mean 1080p with a screen and battery that matches.

SiG · Apr 11, 2020

Governergrimm said:
That's exactly what I mean. If we are saying PS4/X1 levels then I assume 1080p levels. If we say PS4 pro/X1X then I assume 1440-4k. Those screens are not cheap and neither is the power they draw in portable form. Same thing with fps. I guess I am assuming when we say PS4 levels that we mean 1080p with a screen and battery that matches.

Bare in mind resolution alone isn't a good measure of "console power" comparisons as well. A lower resolution just means it takes less processing time (and most likely less watts) to render, but that alone isn't what will help them achieve parity.

Resolution != "power"

Dakhil · Apr 11, 2020

ILikeFeet said:
why would Nintendo choose anything higher than a 1080p screen though? pushing power means they can cram more effects into 1080p. less power overall, cheaper, and at such a small screen size, the differences aren't as perceptable

The only reason I can think of on why Nintendo might want to choose a touchscreen with a higher resolution than 1080p is if Nintendo wants to focus on VR for the "Nintendo Switch 2".

ILikeFeet · Apr 11, 2020

Governergrimm said:
No, I get that. All I'm saying is that portability requires trade offs. A 2021 switch isn't going to be putting out PS4 pro/Xbox 1X at any kind of reasonable price.

It depends on what kind specs we are nailing it down to as well. Are we saying 4k, 1440p, 60fps, 120 fps? If everyone is using a different baseline then we will all have different opinions on possible cost.

I don't think anyone here is expecting Pro/One X levels of performance. PS4-level gpu+better cpu at 1080p is the most people here desire. the fact that DLSS is significantly better, Nintendo can surpass the PS4 simply by rendering at 1/4 of the resolution.

Governergrimm · Apr 11, 2020

SiG said:
Bare in mind resolution alone isn't a good measure of "console power" comparisons as well. A lower resolution just means it takes less processing time (and most likely less watts) to render, but that alone isn't what will help them achieve parity.

Resolution != "power"

I know it isn't. I'm saying that the amount of power and the kind of components is Ill defined. Also when I hear people talk about a switch Pro resolution is one of the improvements listed.

Governergrimm · Apr 11, 2020

ILikeFeet said:
I don't think anyone here is expecting Pro/One X levels of performance. PS4-level gpu+better cpu at 1080p is the most people here desire. the fact that DLSS is significantly better, Nintendo can surpass the PS4 simply by rendering at 1/4 of the resolution.

Sure if the desire is for the pro to hit PS4 levels then yeah I can see it. It'll still be more expensive but yeah THAT I can see.

SiG · Apr 11, 2020

Governergrimm said:
I know it isn't. I'm saying that the amount of power and the kind of components is Ill defined. Also when I hear people talk about a switch Pro resolution is one of the improvements listed.

I think most people talk about how the potential use of Deep Learned Super Sampling can help a Switch successor "cheat" by being able to display higher reconstructed resolutions by using lower base resolutions for rendering and have DLSS to upscale it. Though interestingly, it isn't certain what the end output resolution will be, if they ever do plan to utilize it.

For all we know, the system will just still output 1080p upscaled DLSS from a 720p reconstructed render while using the extra headroom for "next-generation" parity. They might not want to push support for 4k depending on if this exceeds their power threshhold. (Or as some poeple would like to put it, "lol because Nintendo".)

At this point, there are too many unknowns and lots of speculative theoreticals. (Which is good, I guess, since Nintendo has been on info-lockdown ever since those recent lineup leaks.)

Governergrimm · Apr 11, 2020

SiG said:
I think most people talk about how the potential use of Deep Learned Super Sampling can help a Switch successor "cheat" by being able to display higher reconstructed resolutions by using lower base resolutions for rendering and have DLSS to upscale it. Though interestingly, it isn't certain what the end output resolution will be, if they ever do plan to utilize it.

For all we know, the system will just still output 1080p upscaled DLSS from a 720p reconstructed render while using the extra headroom for "next-generation" parity. They might not want to push support for 4k depending on if this exceeds their power threshhold. (Or as some poeple would like to put it, "lol because Nintendo".)

At this point, there are too many unknowns and lots of speculative theoreticals. (Which is good, I guess, since Nintendo has been on info-lockdown ever since those recent lineup leaks.)

Sure that I can agree with. It's just much less fantastic than a good amount of pro speculation that I've seen.

-Le Monde- · Apr 11, 2020

Maple said:
Is anyone else worried that Nintendo will somehow screw things up with their next console? A quick look at their history reveals that they really don't like simply releasing a more powerful version of the same system and adding a number to the end of it. They really want to fundamentally change how people use the system with each and every generation. Gamecube --> Wii --> Wii U. GBA --> DS --> 3DS. They always come up with a new "gimmick" they feel the need push with each new system.

I would love for nothing more than for Nintendo to release a Switch 2 - same design, same concept, fully BC with existing Switch library, with an Ampere based GPU, 8 GB of RAM, powerful ARM cores, OLED display, Bluetooth Audio, etc. But it seems like this goes against what they've been doing as a company for so long. They want to reinvent the wheel with each new generation.

I'm right there with you. The switch is basically the perfect Nintendo system, and I don't want them to change.

With that said, something to keep in mind is that with the switch, Nintendo doesn't have to build a system specifically tied to a gimmick/hook.
With the switch's flexibility, they can innovate(introduce new gimmick), via accessories, like they've been doing with their labo line(vr, driving wheels, robots, fishing controller).

UltraMagnus · Apr 11, 2020

SiG said:
I think most people talk about how the potential use of Deep Learned Super Sampling can help a Switch successor "cheat" by being able to display higher reconstructed resolutions by using lower base resolutions for rendering and have DLSS to upscale it. Though interestingly, it isn't certain what the end output resolution will be, if they ever do plan to utilize it.

For all we know, the system will just still output 1080p upscaled DLSS from a 720p reconstructed render while using the extra headroom for "next-generation" parity. They might not want to push support for 4k depending on if this exceeds their power threshhold. (Or as some poeple would like to put it, "lol because Nintendo".)

At this point, there are too many unknowns and lots of speculative theoreticals. (Which is good, I guess, since Nintendo has been on info-lockdown ever since those recent lineup leaks.)

Rendering at 720p is a waste if you have DLSS in a Switch design setting IMO. You could easily go down to 540p or even less and get a very nice image quality on a 1080p panel for an undocked mode.

UltraMagnus · Apr 11, 2020

ILikeFeet said:
I don't think anyone here is expecting Pro/One X levels of performance. PS4-level gpu+better cpu at 1080p is the most people here desire. the fact that DLSS is significantly better, Nintendo can surpass the PS4 simply by rendering at 1/4 of the resolution.

Wouldn't a PS4 that only has to render 1/4 the pixels be effectively in the end result be as powerful or more powerful than the PS4 Pro?

Pixels are a large component of graphics rendering, they're not some little factor, it impacts everything from system bandwidth to raw teraflop compute performance.

pixelation · Apr 11, 2020

Base XBO.

YolkFolk · Apr 11, 2020

Governergrimm said:
No, I get that. All I'm saying is that portability requires trade offs. A 2021 switch isn't going to be putting out PS4 pro/Xbox 1X at any kind of reasonable price.

It depends on what kind specs we are nailing it down to as well. Are we saying 4k, 1440p, 60fps, 120 fps? If everyone is using a different baseline then we will all have different opinions on possible cost.

Where people are going wrong is the release date. There's little chance of a new Switch release with the power to deliver more than a resolution increase or frame rate increase to current games before the next gen. It's that next gen where the jump will be in terms of matching the PS4 or hopefully the Pro and that jump won't be until 2023 or later.

UltraMagnus · Apr 11, 2020

I dunno about Switch Pro, because I'm not sure if that's happening period, but a Switch 2 in 2022 IMO could be capable of PS5/XB2 ports from the results we're seeing from DLSS 2.0.

1.6 TF undocked/3.3 TF docked with DLSS 2.0 would punch waaaaaay above its weight.

UltraMagnus · Apr 11, 2020

ManCityX said:
Where people are going wrong is the release date. There's little chance of a new Switch release with the power to deliver more than a resolution increase or frame rate increase to current games before the next gen. It's that next gen where the jump will be in terms of matching the PS4 or hopefully the Pro and that jump won't be until 2023 or later.

They can easily destroy a PS4's effective performance with DLSS 2.0 implementation. Easily. Probably right now if they really wanted to, there's no reason for Nintendo to release a new Switch in 2021 though.

PS4 performance is a joke to Nvidia.

YolkFolk · Apr 11, 2020

UltraMagnus said:
They can easily destroy a PS4's effective performance with DLSS 2.0 implementation. Easily. Probably right now if they really wanted to, there's no reason for Nintendo to release a new Switch in 2021 though.

PS4 performance is a joke to Nvidia.

It's just one factor remember.

Battery, game card sizes, game card costs, internal storage, heating, console size, installs etc all need to be at the right level.

ILikeFeet · Apr 11, 2020

UltraMagnus said:
Wouldn't a PS4 that only has to render 1/4 the pixels be effectively in the end result be as powerful or more powerful than the PS4 Pro?

Pixels are a large component of graphics rendering, they're not some little factor, it impacts everything from system bandwidth to raw teraflop compute performance.

whoops. yea I meant PS4P. DLSS is the closest to an actual "secret sauce" that exists

UltraMagnus · Apr 11, 2020

ManCityX said:
It's just one factor remember.

Battery, game card sizes, game card costs, internal storage, heating, console size, installs etc all need to be at the right level.

It's the main factor. Game card sizes get bigger every year and SD Cards with massive storage are dirt cheap.

The fact is most people don't really even know what they're talking about when it comes to mobile processors.

The Apple A12X Bionic is already XBox One tier and that is an old chip (coming up on 1 1/2 years old). The issue is not that this technology can't be made it's that there's no point in having a smartphone or even tablet that powerful.

The only really viable mass market portable solution that needs that kind of power is a game system, and Nintendo is the only main vendor for that and they have no incentive really to release a new Switch in 2021. The current one is sold out and selling just fine.

Even Apple pulled back with the new iPad Pro and gave the processor a more moderate upgrade, because the A12X Bionic was such a monstrously powerful chip for a tablet to begin with, there's no point to going more powerful than that.

YolkFolk · Apr 11, 2020

UltraMagnus said:
It's the main factor. Game card sizes get bigger every year and SD Cards with massive storage are dirt cheap.

The fact is most people don't really even know what they're talking about when it comes to mobile processors.

The Apple A12X Bionic is already XBox One tier and that is an old chip (coming up on 1 1/2 years old). The issue is not that this technology can't be made it's that there's no point in having a smartphone or even tablet that powerful.

The only really viable mass market portable solution that needs that kind of power is a game system, and Nintendo is the only main vendor for that and they have no incentive really to release a new Switch in 2021. The current one is sold out and selling just fine.

Even Apple pulled back with the new iPad Pro and gave the processor a more moderate upgrade, because the A12X Bionic was such a monstrously powerful chip for a tablet to begin with, there's no point to going more powerful than that.

You're arguing points I've never even made.

I'm just saying we won't get a next gen Switch until 2023 or so and not to expect any sort of graphical leap before then if they release another model.

We're also not getting 100gb card games games (COD) before then either.

UltraMagnus · Apr 11, 2020

ManCityX said:
You're arguing points I've never even made.

I'm just saying we won't get a next gen Switch until 2023 or so and not to expect any sort of graphical leap before then if they release another model.

We're also not getting 100gb card games games (COD) before then either.

Depends on what the question is, if the question is what could they do next year if they wanted to ... sure PS4++ performance is probably already possible right now especially if you are talking DLSS technology.

All the PS4 Pro and XBox One X really do is take PS4/XB1 native assets and render them at a higher resolution with a better frame rate in some cases. If you have a mobile chip that can render PS4/XB1 quality, and that already exists, then DLSS implimentation gets you up to PS4 Pro/XB1X level also.

What Nintendo does is ultimately not really even worth worrying about because they do whatever they feel like with no concern and are completely unpredictable. But on a technical level, could Nvidia make a mobile chip that runs say Final Fantasy VII Remake at 1800p docked by next year like the PS4 Pro does currently? Most likely sure.

Just because a game is one size (unoptimized) on PS4/XB1 doesn't really mean that's the only size it can be. Witcher 3 is 50GB on PS4, it's 28GB on the Switch.

Manzoli · Apr 11, 2020

-Le Monde- said:
I'm right there with you. The switch is basically the perfect Nintendo system, and I don't want them to change.

With that said, something to keep in mind is that with the switch, Nintendo doesn't have to build a system specifically tied to a gimmick/hook.
With the switch's flexibility, they can innovate(introduce new gimmick), via accessories, like they've been doing with their labo line(vr, driving wheels, robots, fishing controller).

Well, there's the DS > 3DS jump, where they've basically kept the same design philosophy and just added 3D as a new way to play the same form factor.

Who knows, maybe they'll surprise us and improve the things people want.

I would be satisfied with: more horsepower (keeping 720p handheld and 1080p docked, which I find enough), more battery life, different joy con form factor (more like the split pad pro, with a DPad), with good hardware of course, a good LCD IPS panel, and, to top it all, Bluetooth audio support.

jdstorm · Apr 11, 2020

If we are speculating on Switch 2 and Launch games I think it will be something like this

Release Date: Holiday 2022 (say September)

Launch games: Aka September 2022-December 2022

Mario Kart 9
Smash Bros Ultimate Deluxe
Splatoon 3
Next mainline Pokémon (Open World)
New IP
Some sort of Single Player action game (Astral 2 Chainz?)

Sander VF · Apr 11, 2020

jdstorm said:
If we are speculating on Switch 2 and Launch games I think it will be something like this

Release Date: Holiday 2022 (say September)

Launch games: Aka September 2022-December 2022

Mario Kart 9
Smash Bros Ultimate Deluxe
Splatoon 3
Next mainline Pokémon (Open World)
New IP
Some sort of Single Player action game (Astral 2 Chainz?)

I fully expect Switch 2 to be 100 % backwards compatible so the bolded isn't something I expect in any capacity.

jdstorm · Apr 11, 2020

Sander VF said:
I fully expect Switch 2 to be 100 % backwards compatible so the bolded isn't something I expect in any capacity.

I agree with you on the backwards comparability. I just don't see how Nintendo don't release a retail version of the complete Smash Ultimate (all fighter pass characters, higher resolution ect) at the start of next gen.

Smash Ultimate at this point is too big to make a sequel too. May as well just port the Deluxe version and keep adding fighters.

Core players would buy it for some extra mode (single player?) and casuals would buy it because it was the Defult Smash Bros game

BlackTangMaster · Apr 12, 2020

A 5 nm Switch in 2022/2023 may be too late when current Switch is already having trouble running first-party games at native resolutions. Each Nintendo big IP has already seen a release. An early 2021 upgrade could offer new iterations to these IPs with better graphics aimed at early birds with 4 yo switch and those who didn't have it yet.

It would be to hard to find a business case for new games aimed at 2023 with such a large technical gap if they could not run on the 70 million Switch 2017/2019 (instead of running on 15/30 million Switch 2021). 2017 Switch could not run games at docked mod in handeld. No interest in making next-gen games (mostly portages) for 2 to 5 million consoles. From strategic point of view, it's the perfect time to release a console capable of PS4 performances: exclusive games from a few publishers, and numerous ports of actual gen. Launch models would run these games at docked frequencies in portable. We would have sequels to big IPs that would bring new comers and make firsts ones buying a new model.

From a technical point of view, latest rumors are saying that Nvidia new GPUs and Tegra Soc have been tapped out on Samsung 8nm LPP (maybe LPU with density close to Samsung 7 LPP ). There is a high probability that the next switch chip (2021 or even 2022/2023 one) is the one planned on this mature node. It would have the same architecture (ampere?) as the last desktop GPUs which can later be brought to 6 or 5nm process in 2024 on mature nodes.

My predictions for a Switch in 2021 without "2" in its name:

New 7W in portable to 15W docked Nvidia Tegra SOC on Samsung 8 LPU

12 cores ARM CPU
8 high power A76/A77
4 low power A55@1-2Ghz (able to run OG games ?)
512 gpu cores @1.4Ghz docked at 1.4Tflop (>PS4 level) with tensor core, VRS and DLSS
2x4Gb or 2x8Gb LPDDR5 ram modules
64 or 128 Gb eUFS memory
720p LCD screen with 'freesync' (don't know if HDR is possible)
Same form factor with bigger screen. Maybe magnesium or aluminium frame.

It's the more probable chip for the next three years because of 5/6nm being still in risk production and not rentable enough. AMD's best 7nm SOCs offers 1.8 tflops at 15W and my prediction matches with a die shrunk Xavier SOC in 8 nm (more than double in density and the same size as a tegra X1) without Carmel cores.

These designs are planned several years in advance and games planned for next-gen consoles cannot be delayed indefinitely (MH switch and FFVIIR switch?) even in the current context.

Why put a cooler on Switch lite if it was not intended to run it on docked frequencies? That would have no interest with a successor in 2023. Why calling a defect 15W Xavier chip 'NX'?

Zedark · Apr 12, 2020

jdstorm said:
If we are speculating on Switch 2 and Launch games I think it will be something like this

Release Date: Holiday 2022 (say September)

Launch games: Aka September 2022-December 2022

Mario Kart 9
Smash Bros Ultimate Deluxe
Splatoon 3
Next mainline Pokémon (Open World)
New IP
Some sort of Single Player action game (Astral 2 Chainz?)

That sounds closer to a launch year line-up than a launch quarter(-ish) line-up to me. But it would be a great start for a Switch 2!

BlackTangMaster said:
A 5 nm Switch in 2022/2023 may be too late when current Switch is already having trouble running first-party games at native resolutions. Each Nintendo big IP has already seen a release. An early 2021 upgrade could offer new iterations to these IPs with better graphics aimed at early birds with 4 yo switch and those who didn't have it yet.

It would be to hard to find a business case for new games aimed at 2023 with such a large technical gap if they could not run on the 70 million Switch 2017/2019 (instead of running on 15/30 million Switch 2021). 2017 Switch could not run games at docked mod in handeld. No interest in making next-gen games (mostly portages) for 2 to 5 million consoles. From strategic point of view, it's the perfect time to release a console capable of PS4 performances: exclusive games from a few publishers, and numerous ports of actual gen. Launch models would run these games at docked frequencies in portable. We would have sequels to big IPs that would bring new comers and make firsts ones buying a new model.

From a technical point of view, latest rumors are saying that Nvidia new GPUs and Tegra Soc have been tapped out on Samsung 8nm LPP (maybe LPU with density close to Samsung 7 LPP ). There is a high probability that the next switch chip (2021 or even 2022/2023 one) is the one planned on this mature node. It would have the same architecture (ampere?) as the last desktop GPUs which can later be brought to 6 or 5nm process in 2024 on mature nodes.

My predictions for a Switch in 2021 without "2" in its name:

New 7W in portable to 15W docked Nvidia Tegra SOC on Samsung 8 LPU

12 cores ARM CPU

8 high power A76/A77

4 low power A55@1-2Ghz (able to run OG games ?)

512 gpu cores @1.4Ghz docked at 1.4Tflop (>PS4 level) with tensor core, VRS and DLSS

2x4Gb or 2x8Gb LPDDR5 ram modules

64 or 128 Gb eUFS memory

720p LCD screen with 'freesync' (don't know if HDR is possible)

Same form factor with bigger screen. Maybe magnesium or aluminium frame.

It's the more probable chip for the next three years because of 5/6nm being still in risk production and not rentable enough. AMD's best 7nm SOCs offers 1.8 tflops at 15W and my prediction matches with a die shrunk Xavier SOC in 8 nm (more than double in density and the same size as a tegra X1) without Carmel cores.

These designs are planned several years in advance and games planned for next-gen consoles cannot be delayed indefinitely (MH switch and FFVIIR switch?) even in the current context.

Why put a cooler on Switch lite if it was not intended to run it on docked frequencies? That would have no interest with a successor in 2023. Why calling a defect 15W Xavier chip 'NX'?

While it's obviously true that the next chip design for a Switch 2 releasing 2023 won't be finished as late as 2022, I don't think the 2020 technology barrier indicated by several in this thread is at all unlikely: the technology in the Switch is 2014 technology for a 2017 machine, after all. TSMC is aiming to bring 5nm to the mass market this year, so it will be 2020 technology as well (of course, it depends a bit on what impact the COVID-19 virus has on their progress). A 5nm machine in 2023 is not an unlikely proposition imo.

Now, I don't agree that Switch cannot last until 2023. Your argument as I understand it has three components: firstly, the Switch struggles to run games already, and it can't last until 2023 without some sort of update; secondly, Nintendo won't want to do next-gen exclusives at launch on a machine that has only a few million users; thirdly, most major IP already have had an iteration on the Switch. I will discuss these points below:

- Struggling to run games : It's true! SMO, Splatoon 2, BOTW, and more all failed to hit 720p/1080p target resolutions. Some less optimised games go even lower, for example Yoshi's Crafted World and Xenoblade Chronicles 2. It's clear that Nintendo is having some issue running their software on the system. While a 2021 Switch "Pro" does have a clear sales opportunity, just like the PS4Pro and the XB1X have, it doesn't solve the problem of struggling to run the system. And even then, the Pro/X systems have tended to sell only about 1/4 (or 1/5?) of the total system sales aligned with the current base system sales, so we are looking at a rather small part of the userbase buying into those systems in the grand scheme of things. It's a useful boost for sure - but it's not a business practice that allows you to lower your demands for the base versions, and as such, development on the main Switch will continue to struggle (although I'm sure the next line of software will show good improvement on the previous line of software: devs always learn better ways to make use of the system over time). An intermediate system does not alleviate dev concerns in a significant way, therefore.

- Most major IP are already present : True. I think 2D Mario will see a new entry still, and games like Style Savvy and Tomodachi could see a new entry as well. But for the most part, the major first party IP are present on the system. However, that doesn't mean that the existing IP which are going to get a new entry will not be able to sell the system to a major degree. We know that FIFA and Call of Duty sell systems to a major degree each and every holiday season. Why? Well, one part of it is that each year, new people reach the age where they become interested in these games, and what better way to jump in than with the newest iteration? But there's a more relevant effect of software to be considered imo: that of the critical mass. Switch is widely lauded for its large amount of strong software. Indeed, software sales for its already released games are incredibly strong. The system has a hugely appealing back catalog, which only grows in strength due to the staying power of most of their games. In that situation, the addition of big games, either from existing franchise or new ones, will do a lot for the hardware as it provides an extra jolt to get people to take a jump on the system. In my opinion the Switch can hold a strong baseline of sales, and benefit from strong boosts whenever major software launches. In fact, the system's secret weapon is Pokémon, which releases usually about 3-4 times on a system, and sells tons of hardware whenever it releases. That in itself gives a strong push towards the idea that the system can survive for quite a bit longer than just 2021 - Pokémon really is that big, and 2021 and 2023 will probably see new entries release on the system (one remake and one new gen game I think).

- The sales argument for non-exclusives : Nintendo software is the most suitable software to be launched in the early months of a system's launch. This is of course due to their extreme evergreen status, which we have seen demonstrated with the Switch itself: before the system hit 10M, it had Splatoon 2, BOTW, SMO, and MK8D, each of which are major Nintendo IP. All of those games have gone on to sell a ton of units in the long run, and were not impeded by their launch early on in the system's launch. This is the Switch 2's business proposal: they can use the power of their software to sell their new system without noticeably impacting the lifetime sales potential of their games. It's actually a strong business point: it helps to solidify their future system by propping the system up with major exclusives, all the while being able to rely on their software selling a lot of units regardless. I therefore do not see a problem in business terms for not launching with a suite of next-gen exclusives during the launch year of the Switch 2.

To conclude, then, I do not agree with the idea that a 2023 system is too late, and I think 2020 technology is roughly the ballpark for new technology that can be expected in a Switch 2, which concretely suggests 5nm process node and ARM A78 or A77 CPUs, along with fast internal storage of 1 GB/s or higher read speeds.

BlackTangMaster · Apr 12, 2020

Zedark said:
That sounds closer to a launch year line-up than a launch quarter(-ish) line-up to me. But it would be a great start for a Switch 2!

While it's obviously true that the next chip design for a Switch 2 releasing 2023 won't be finished as late as 2022, I don't think the 2020 technology barrier indicated by several in this thread is at all unlikely: the technology in the Switch is 2014 technology for a 2017 machine, after all. TSMC is aiming to bring 5nm to the mass market this year, so it will be 2020 technology as well (of course, it depends a bit on what impact the COVID-19 virus has on their progress). A 5nm machine in 2023 is not an unlikely proposition imo.

Now, I don't agree that Switch cannot last until 2023. Your argument as I understand it has three components: firstly, the Switch struggles to run games already, and it can't last until 2023 without some sort of update; secondly, Nintendo won't want to do next-gen exclusives at launch on a machine that has only a few million users; thirdly, most major IP already have had an iteration on the Switch. I will discuss these points below:

- Struggling to run games : It's true! SMO, Splatoon 2, BOTW, and more all failed to hit 720p/1080p target resolutions. Some less optimised games go even lower, for example Yoshi's Crafted World and Xenoblade Chronicles 2. It's clear that Nintendo is having some issue running their software on the system. While a 2021 Switch "Pro" does have a clear sales opportunity, just like the PS4Pro and the XB1X have, it doesn't solve the problem of struggling to run the system. And even then, the Pro/X systems have tended to sell only about 1/4 (or 1/5?) of the total system sales aligned with the current base system sales, so we are looking at a rather small part of the userbase buying into those systems in the grand scheme of things. It's a useful boost for sure - but it's not a business practice that allows you to lower your demands for the base versions, and as such, development on the main Switch will continue to struggle (although I'm sure the next line of software will show good improvement on the previous line of software: devs always learn better ways to make use of the system over time). An intermediate system does not alleviate dev concerns in a significant way, therefore.

- Most major IP are already present : True. I think 2D Mario will see a new entry still, and games like Style Savvy and Tomodachi could see a new entry as well. But for the most part, the major first party IP are present on the system. However, that doesn't mean that the existing IP which are going to get a new entry will not be able to sell the system to a major degree. We know that FIFA and Call of Duty sell systems to a major degree each and every holiday season. Why? Well, one part of it is that each year, new people reach the age where they become interested in these games, and what better way to jump in than with the newest iteration? But there's a more relevant effect of software to be considered imo: that of the critical mass. Switch is widely lauded for its large amount of strong software. Indeed, software sales for its already released games are incredibly strong. The system has a hugely appealing back catalog, which only grows in strength due to the staying power of most of their games. In that situation, the addition of big games, either from existing franchise or new ones, will do a lot for the hardware as it provides an extra jolt to get people to take a jump on the system. In my opinion the Switch can hold a strong baseline of sales, and benefit from strong boosts whenever major software launches. In fact, the system's secret weapon is Pokémon, which releases usually about 3-4 times on a system, and sells tons of hardware whenever it releases. That in itself gives a strong push towards the idea that the system can survive for quite a bit longer than just 2021 - Pokémon really is that big, and 2021 and 2023 will probably see new entries release on the system (one remake and one new gen game I think).

- The sales argument for non-exclusives : Nintendo software is the most suitable software to be launched in the early months of a system's launch. This is of course due to their extreme evergreen status, which we have seen demonstrated with the Switch itself: before the system hit 10M, it had Splatoon 2, BOTW, SMO, and MK8D, each of which are major Nintendo IP. All of those games have gone on to sell a ton of units in the long run, and were not impeded by their launch early on in the system's launch. This is the Switch 2's business proposal: they can use the power of their software to sell their new system without noticeably impacting the lifetime sales potential of their games. It's actually a strong business point: it helps to solidify their future system by propping the system up with major exclusives, all the while being able to rely on their software selling a lot of units regardless. I therefore do not see a problem in business terms for not launching with a suite of next-gen exclusives during the launch year of the Switch 2.

To conclude, then, I do not agree with the idea that a 2023 system is too late, and I think 2020 technology is roughly the ballpark for new technology that can be expected in a Switch 2, which concretely suggests 5nm process node and ARM A78 or A77 CPUs, along with fast internal storage of 1 GB/s or higher read speeds.

The key point of a Switch 2021 is that it would allow third party games to be able to run on OG switch with docked clocks at the cost of battery life. This is not possible with >1 Tflop undocked switch 2023.

Waiting until 2023 would bring us back to square one with publishers not wanting to port heavy PS4 and X1 games (the sole 'real' problem of OG switch) to such a small amount of 2023 or 2024 switches. It splits the user base. And I am not even talking about PS5 games that won't be bottlenecked by netbook CPUs and outdated GPU at release. There will be a business case for a game aimed for a 2023 switch with 20 million 2021 switches sold, not with 0 new switches and 80 million incompatible OG.

In the end, it will be difficult to attract new consumers with big IPs already out (some with a sequel like BOTW2 and probably an new 3D Mario) if the console remains the same until 2023. Those who want to play Zelda BOTW2 have already bought the console in 2017. Once all the evergreen are released, only a price drop or a more powerful console will attract those who could not afford or those who found the console too weak.

In any case, lowering the price of OG (and cut production) will free the niche of $ 299 / € 329 for a more powerful Switch. History has already shown that people will prefer the most complete model (lite vs OG). This intermediate model will be essential in order not to separate the user base in 2023/2024. Game developments of big exclusives or heavy current gen port can be launched now and, in the worst case, run on 60 million OG console rather than speculating on the success of a future model which will no longer benefit from the scope of big IPs of OG switch. The 2023 gen Zelda or Mario game may not be as good as their predecessors at selling consoles.

jdstorm · Apr 13, 2020

Zedark said:
That sounds closer to a launch year line-up than a launch quarter(-ish) line-up to me. But it would be a great start for a Switch 2!

While it's obviously true that the next chip design for a Switch 2 releasing 2023 won't be finished as late as 2022, I don't think the 2020 technology barrier indicated by several in this thread is at all unlikely: the technology in the Switch is 2014 technology for a 2017 machine, after all. TSMC is aiming to bring 5nm to the mass market this year, so it will be 2020 technology as well (of course, it depends a bit on what impact the COVID-19 virus has on their progress). A 5nm machine in 2023 is not an unlikely proposition imo.

Now, I don't agree that Switch cannot last until 2023. Your argument as I understand it has three components: firstly, the Switch struggles to run games already, and it can't last until 2023 without some sort of update; secondly, Nintendo won't want to do next-gen exclusives at launch on a machine that has only a few million users; thirdly, most major IP already have had an iteration on the Switch. I will discuss these points below:

- Struggling to run games : It's true! SMO, Splatoon 2, BOTW, and more all failed to hit 720p/1080p target resolutions. Some less optimised games go even lower, for example Yoshi's Crafted World and Xenoblade Chronicles 2. It's clear that Nintendo is having some issue running their software on the system. While a 2021 Switch "Pro" does have a clear sales opportunity, just like the PS4Pro and the XB1X have, it doesn't solve the problem of struggling to run the system. And even then, the Pro/X systems have tended to sell only about 1/4 (or 1/5?) of the total system sales aligned with the current base system sales, so we are looking at a rather small part of the userbase buying into those systems in the grand scheme of things. It's a useful boost for sure - but it's not a business practice that allows you to lower your demands for the base versions, and as such, development on the main Switch will continue to struggle (although I'm sure the next line of software will show good improvement on the previous line of software: devs always learn better ways to make use of the system over time). An intermediate system does not alleviate dev concerns in a significant way, therefore.

- Most major IP are already present : True. I think 2D Mario will see a new entry still, and games like Style Savvy and Tomodachi could see a new entry as well. But for the most part, the major first party IP are present on the system. However, that doesn't mean that the existing IP which are going to get a new entry will not be able to sell the system to a major degree. We know that FIFA and Call of Duty sell systems to a major degree each and every holiday season. Why? Well, one part of it is that each year, new people reach the age where they become interested in these games, and what better way to jump in than with the newest iteration? But there's a more relevant effect of software to be considered imo: that of the critical mass. Switch is widely lauded for its large amount of strong software. Indeed, software sales for its already released games are incredibly strong. The system has a hugely appealing back catalog, which only grows in strength due to the staying power of most of their games. In that situation, the addition of big games, either from existing franchise or new ones, will do a lot for the hardware as it provides an extra jolt to get people to take a jump on the system. In my opinion the Switch can hold a strong baseline of sales, and benefit from strong boosts whenever major software launches. In fact, the system's secret weapon is Pokémon, which releases usually about 3-4 times on a system, and sells tons of hardware whenever it releases. That in itself gives a strong push towards the idea that the system can survive for quite a bit longer than just 2021 - Pokémon really is that big, and 2021 and 2023 will probably see new entries release on the system (one remake and one new gen game I think).

- The sales argument for non-exclusives : Nintendo software is the most suitable software to be launched in the early months of a system's launch. This is of course due to their extreme evergreen status, which we have seen demonstrated with the Switch itself: before the system hit 10M, it had Splatoon 2, BOTW, SMO, and MK8D, each of which are major Nintendo IP. All of those games have gone on to sell a ton of units in the long run, and were not impeded by their launch early on in the system's launch. This is the Switch 2's business proposal: they can use the power of their software to sell their new system without noticeably impacting the lifetime sales potential of their games. It's actually a strong business point: it helps to solidify their future system by propping the system up with major exclusives, all the while being able to rely on their software selling a lot of units regardless. I therefore do not see a problem in business terms for not launching with a suite of next-gen exclusives during the launch year of the Switch 2.

To conclude, then, I do not agree with the idea that a 2023 system is too late, and I think 2020 technology is roughly the ballpark for new technology that can be expected in a Switch 2, which concretely suggests 5nm process node and ARM A78 or A77 CPUs, along with fast internal storage of 1 GB/s or higher read speeds.

See since they are evergreens might as well stack the launch quarter. Nintendo has been pretty backhalf heavy anyway

Mr. Pointy · Apr 13, 2020

One thing I'd like to know is how fast the next Switch's internal memory/carts/SD card support is going to be for loading games. The Switch only supports UHS-I SD cards, which are supposed to max out at 104MB/sec. The Switch's internal memory, cartridges and SD card read speed aren't pulling data that fast - the SD card transfer rate on Switch seems to top out at 80MB/sec. I know the CPU speed limits that, as the temporary 1.8GHz clock profile is only used for loading. I don't know what other bottlenecks there are.

UHS-II micro SD cards are available, and they theoretically max out at 312MB/sec, but they are expensive (Sandisk have a 128GB UHS-II card that has a 275MB/sec transfer rate but it's dearer than their 512GB UHS-I card). But we can presume that demand will bring prices of those cards down in the next 2 years.

And UHS-III isn't out yet.

If we presume 3-4x faster load times for games on the 'Switch 2', using UHS-II as a basis, I could guess that load times are still going to be 3-4x longer than games on the PS5 and XSX.

Jesb · Apr 13, 2020

They need a high quality screen. I think that should be a priority. They definitely need bc and a good battery life. I expect PS4 level at 1080p as a minimum.

UltraMagnus · Apr 13, 2020

Maple said:
Is anyone else worried that Nintendo will somehow screw things up with their next console? A quick look at their history reveals that they really don't like simply releasing a more powerful version of the same system and adding a number to the end of it. They really want to fundamentally change how people use the system with each and every generation. Gamecube --> Wii --> Wii U. GBA --> DS --> 3DS. They always come up with a new "gimmick" they feel the need push with each new system.

I would love for nothing more than for Nintendo to release a Switch 2 - same design, same concept, fully BC with existing Switch library, with an Ampere based GPU, 8 GB of RAM, powerful ARM cores, OLED display, Bluetooth Audio, etc. But it seems like this goes against what they've been doing as a company for so long. They want to reinvent the wheel with each new generation.

I think the "gimmick" for Switch 2 could be an *optional* dedicated VR helmet (plastic, no screen, so cheap) with Joycons that have more sensory functions and a headset that can mimic things like wind blowing in your face.

Seems "Nintendo-ey" enough and they could implement it into games like Mario Kart (already has a VR mode for arcades) and Pilotwings and Metroid Prime. I think there was talk of VR usage at Nintendo's Universal Studios attractions too so maybe Miyamoto and co. want to bring some of what they learned designing that to the next system.

Labo VR I think was the test run for a real VR experience on Switch 2. Nintendo tends to do this in that they don't throw ideas away.

ILikeFeet · Apr 13, 2020

Mr. Pointy said:
One thing I'd like to know is how fast the next Switch's internal memory/carts/SD card support is going to be for loading games. The Switch only supports UHS-I SD cards, which are supposed to max out at 104MB/sec. The Switch's internal memory, cartridges and SD card read speed aren't pulling data that fast - the SD card transfer rate on Switch seems to top out at 80MB/sec. I know the CPU speed limits that, as the temporary 1.8GHz clock profile is only used for loading. I don't know what other bottlenecks there are.

UHS-II micro SD cards are available, and they theoretically max out at 312MB/sec, but they are expensive (Sandisk have a 128GB UHS-II card that has a 275MB/sec transfer rate but it's dearer than their 512GB UHS-I card). But we can presume that demand will bring prices of those cards down in the next 2 years.

And UHS-III isn't out yet.

If we presume 3-4x faster load times for games on the 'Switch 2', using UHS-II as a basis, I could guess that load times are still going to be 3-4x longer than games on the PS5 and XSX.

Internal memory is an easy solution with UFS (peaks at 2100MB/s) or an home-based solution like apple uses.

With external storage, Nintendo will either have to do what Microsoft does and use an expensive format (like SD Express (985MB/s) or UFS) or just swap objects off of an SD card onto internal memory

BlackTangMaster · Apr 13, 2020

ILikeFeet said:
Internal memory is an easy solution with UFS (peaks at 2100MB/s) or an home-based solution like apple uses.

With external storage, Nintendo will either have to do what Microsoft does and use an expensive format (like SD Express (985MB/s) or UFS) or just swap objects off of an SD card onto internal memory

A new model could even have 12 (even 16 ?) gb ram with less expensive memory. Most first-party games take less than 15 GB. Switch was released with 4 Gb LPDDR4 and 32 eMMC when the Galaxy S7 had 4 Gb LPDDR4 and 32 Gb UFS 2.0.

A new switch could easily have Galaxy S10's 8 to 12 Gb LPDDR4 and 128 Gb UFS 2.1. Even with LPDDR5 and UFS 3.0 if it comes out after 2022. Poco X2 sells for less than $ 300 with 8 Gb LPDDR4 and 256 Gb UFS 2.1 with HDR 120 Hz LCD and lots of useless things compared to a switch like cameras.

ILikeFeet · Apr 13, 2020

BlackTangMaster said:
A new model could even have 12 (even 16 ?) gb ram with less expensive memory. Most first-party games take less than 15 GB. Switch was released with 4 Gb LPDDR4 and 32 eMMC when the Galaxy S7 had 4 Gb LPDDR4 and 32 Gb UFS 2.0.

A new switch could easily have Galaxy S10's 8 to 12 Gb LPDDR4 and 128 Gb UFS 2.1. Even with LPDDR5 and UFS 3.0 if it comes out after 2022. Poco X2 sells for less than $ 300 with 8 Gb LPDDR4 and 256 Gb UFS 2.1 with HDR 120 Hz LCD and lots of useless things compared to a switch like cameras.

They aren't gonna go as high as 16GB of ram. The most we could possibly see is 12GB and that's stretching it.

BlackTangMaster · Apr 13, 2020

You're probably right for 12Gb. Normally 8gb is a given since we will probably have 2 ram chip for bandwidth reason and that samsung should probably no longer produce 2 Gb ram ?

gamer forever · Apr 13, 2020

Knowing nintendo it will be a ps3 1.5 level, but with a 1080p screen and possibly up to 4k tv output (but most games being 1080p).

Nintendo aim to make a lot of profit off their hardware, so no chance of advanced tech, but then we don't need it with ps5/ xbox series x being around.

Deleted member 18161 · Apr 13, 2020

UltraMagnus said:
I think internally though the unthinkable (at least probably in Nintendo's eyes) poor launches of the Wii U and 3DS after the phenomenons that were the Wii and DS still is a memory seared into Nintendo's conciousness.

The big boogeyman now is not launching a successful console ... it's can they follow up a successful console without problems.

Mario Kart 9 just removes a lot of that worry and headache. Also it's a more straight forward game to develop ... a Zelda game or a Mario 3D game ... who knows, shit can go sideways in development and require a delay but you got a hardware platform to launch and "oops" is not a good enough answer.

MK8 DX basically continues to sell like a brand new game, I just think Nintendo is gonna say "nah" to another one on the current Switch. Same with Splatoon 3, I think that is also ear marked for Switch 2. The two of those in the first 3-4 months is the type of 1-2 punch you want launching a crucial platform.

Having a smooth first 6 months is just so crucial, it saves you so much grief in the long run, when you have a launch get away from you, you're really scrambling hard sometimes for the rest of the generation really. MK9 is a perfect fail safe for Switch 2.

I agree with you 100%. As big has having BotW at launch and SMO in fall, MK8D and Splatoon 2 were just as important for Switch's blistering first years sales start. The importance of hardware sales momentum in years 1-3 cannot be understated when thinking of new console launches and that momentum is driven by exciting, exclusive software. Nintendo have got a lot of internal data that shows this based on their own failed consoles.

Were I Nintendo I would put out another DLC pack for MK8D next Spring (more characters + tracks from other series basically teasing the long rumoured conversion to Nintendo Kart without changing the name 😋) while keeping MK9 as part of the one-two punch for the launch of the next gen Switch.

I would also keep the next new 3D Mario for the consoles first Holiday season especially considering they're apparently doing the 64/Sunshine/Galaxy collection and 3D World Remaster in the coming year. They have to be careful with series fatigue.

They can repeat the trick with Switch 2 by releasing a new Zelda at launch (BotW 2 will probably be pushed to late 2022 meaning a Switch 2 version at launch would fit nicely), MK9 in Spring 2023, Splatoon 3 in Summer 2023 then a new 3D Mario for Holiday season 2023. It really isn't rocket science when you have the kind of exclusive IP Nintendo has. The difficulty is obviously lining up the software timing.

Based on how tech has progressed since the Switch, how powerful could a Switch Pro in early 2021 actually be?

Attempted to circumvent ban with an alt-account

DF Deet Master

DF Deet Master

DF Deet Master

DF Deet Master

DF Deet Master

DF Deet Master

DF Deet Master

DF Deet Master

user requested account closure