Support NeoGAF

[Dorfdad]

The Switch 2 has been a massive success relative to expectations for its first year, but because the console was a known entity long before release, its hardware limitations were largely baked in. DLSS has certainly helped, but I am wondering if there is any untapped potential left to exploit going forward.

Frame Gen doesn't seem possible with the current hardware. While we will likely see incremental DLSS refinements, does anyone believe we can squeeze out a 30% performance improvement over the rest of the system's lifespan, or are we already at the pinnacle—meaning only a hardware revision could push things further?

Perhaps a Switch 2 OLED model would include a "Pro" spec bump, utilizing the new screen to finally hit those 4K/60 targets?

what's your take on the switch 2 moving forward!? Honestly interested!

 

[OverHeat]

Nope

 

[Witchilich]

Secret Sauce again?

 

[FoxMcChief]

Nope. None. Switch 2 peaked 3 years ago.

 

[blacktout]

There was talk that Nintendo might eventually be able to free up some of the RAM and CPU power currently reserved for the OS, especially if they allow devs to deactivate GameChat functionality while their game is running. Not sure when or if that will actually happen.

There are also definitely going to be improvements as developers learn the nuances of the hardware and various game engines implement optimizations tailored specifically to the Switch 2.

 

[kevboard]

uh... not really.

the times of devs learning hardware over years of optimising stuff for it are long over.

these days you basically see what a system can do within the first year or so

 

[Tarnpanzer]

these days you basically see what a system can do within the first year or so

If you look at very good ports like SW:Outlaws, Nintendos First Party Titles really fall a bit short in comparison from the technical site. I hope all year 2 titles and beyond use proper DLSS.

I also really hope they somehow make VRR in docked mode possible.

 

[ThisIsMyDog]

it's a underpowered tablet, so the same as original Switch, don't expect much.

 

[RoboFu]

There's all the head room ..

 

[Bonfires Down]

No. Get the PS6 Portable if you need more power.

 

[PeteBull]

It has entry lvl 2021 nvidia tech, and its launch was pushed back solid couple years coz og switch was selling like crazy, we likely saw max or close to max of its capabilities already, so in future games something gonna have to give, resolution, framerate or graphic fidelity, normally when we talking stationary console hardware u can only have 2 out of 3, in case of switch2 since its so weak(and has handheld hardware) we will likely only gonna get 2 out of 3 too, sometimes maybe only 1 out of 3...

Next jrpg from monolith soft likely gonna be final litmus test, if that doesnt look impressive likely nothing will(and keep in mind it will likely have terribly bad image quality and framerate, same like their xenoblade chronicles 3 had on og switch)
For comparision, thats how SW: O looks and runs on switch2


On super weak(but still much stronger) 3050 it looks visibly better/at higher settings and runs way smoother too, timestamped so u can compare directly:


And just for reference, look where desktop rtx 3050 8gigs version stacks up among modern graphic cards (it launched jan 2022 on 8nm, 250$ msrp):

NVIDIA GeForce RTX 3050 8 GB Specs

NVIDIA GA106, 1777 MHz, 2560 Cores, 80 TMUs, 32 ROPs, 8192 MB GDDR6, 1750 MHz, 128 bit

www.techpowerup.com

Again for comparision, 16gigs 9060xt aka ps5pr0 equivalent is roughly at 257% perf lvl in games(in other worlds 2,5x stronger), nuff said...

True problematic times for switch2 graphical prowess(or rather lack there of) will start durning x-gen so end of 2027 and onwards, by that time many multiplats either gonna have to be downgraded to hell and back to run on it or devs gonna decide its not worth their effort and simply not port the game at all...
So u got about 1,5 years left, after that brace for impact

 

[DarkShadowXI]

I already set my expectations before launch.

Much easier to accept that you're better off going elsewhere for 3rd party. But if you like Nintendo, what we get is what we get.

If this is your only system, then 3rd party will just be...okay.

 

[Cancer]

Even in 8K and 240 frames per second, Metroid Prime 4 will always be the child that should have remained hidden in the basement.

 

[SrBelmont]

If you compare the OG Switch with the Switch 2, it's clear Nintendo is in a much better position

 

[S0ULZB0URNE]

Absolutely.
We haven't even seen Nintendo mess with DLSS yet.

 

[Notorious_Roy]

No. Get the PS6 Portable if you need more power.

I'll stop by Toys R Us later today to pick it up.

 

[Hollowpoint5557]

I think its impressive already for what it is. I didnt buy one thinking it was going to be a graphical powerhouse.

 

[Trogdor1123]

30% would be crazy….

 

[MCH2024]

The Switch 2 has been a massive success relative to expectations for its first year, but because the console was a known entity long before release, its hardware limitations were largely baked in. DLSS has certainly helped, but I am wondering if there is any untapped potential left to exploit going forward.

Frame Gen doesn't seem possible with the current hardware. While we will likely see incremental DLSS refinements, does anyone believe we can squeeze out a 30% performance improvement over the rest of the system's lifespan, or are we already at the pinnacle—meaning only a hardware revision could push things further?

Perhaps a Switch 2 OLED model would include a "Pro" spec bump, utilizing the new screen to finally hit those 4K/60 targets?

what's your take on the switch 2 moving forward!? Honestly interested!

Yes the T239 Max GPU Clock is 1.4 GHz. So All existing models have that headroom.

Putting that aside. T239 is on SEC 10nm++ (7LPH/8NP). A shrink to 2-3nm could easily increase performance 2-3x.

 

[MCH2024]

It has entry lvl 2021 nvidia tech

Equivalent or superior to 2025 AMD Tech. And?

 

[Sebastian1295]

The switch 2 specs are below average by modern pc and console standards. Of course there's no headroom.

 

[Cumulus]

Every gen the same "but but but secret sauce" bullshit from the "fans"......

 

[ChoosableOne]

There's all the head room ..

The Head:

 

[ReyBrujo]

Nintendo isn't likely to release a "Pro" version. Yes, they will release an oled version but it will pretty much have the same processor just as the Switch v1/v2, lite and oled. But then again did the v1 have 30% performance jump?
Now, for the lite they will probably get a more optimized version of the same chip but more powerful? Nah.

 

[hardwood007]

There's talk that the Samsung 8nm node that it's on is a dead end node. Is there really no way for a die shrink in the future? The battery life is abysmal

 

[uncreativename]

My take is that with enough creativity, talent and money you can do amazing things even within tight constraints. So, basically unlimited headroom.

 

[NeverYouMind]

Not in the way you think. It is a pretty decent piece of kit for the price and if developers optimize for Switch 2 it can definitely punch above its weight. The only real issue is Unreal Engine 5 being the go to engine for most games and developers not stripping it of unneeded features that add a lot of overhead. This problem is not exclusive to Switch 2 either.

 

[DonkeyPunchJr]

once they enable the secret discrete GPU, it will be on par with PS5 if not PS5 Pro

 

[SoloCamo]

They will obviously squeeze a bit more out of it as they gain more experience with it, but that's every console ever produced so.... don't expect miracles.

 

[UltimaKilo]

There's talk that the Samsung 8nm node that it's on is a dead end node. Is there really no way for a die shrink in the future? The battery life is abysmal

Of course, eventually they will get down to 6 or maybe 4nm. More per wafer = savings. If they can get the weight down further, that could save them some. It will lead to a more efficient console with maybe an extra hour of battery life.

 

[daclynk]

i mean Nintendo did TOTK on the switch, which was after BOTW on WiiU, with the tech they have access to. their DLSS like tech, 3D audio, RTX, HDR2 tech, 4K and 120fps. Im not worried about what gameplay design evolution or gameplay tech they will come up with.
OHH and their GAMES are going to LOOK AMAZING now, going by what we see them doing with Star Fox and thats a Remaster.

 

[Vaquilla]

Systems almost never show their absolute best in their first year, as engines and such improve over time.
There's been nearly nothing so far built from the ground up to push the hardware, as Nintendo's first party games have been very technically conservative and third party stuff is pretty much entirely ports.

The best is almost certainly still to come.

 

[PeteBull]

Equivalent or superior to 2025 AMD Tech. And?

We can agree best case scenario it trades blows with series s, and we all know what will happen to that console durning xgen period aka in 2028 and onwards, with base ps5 that has much bigger instalbase and is almost 3x stronger devs will support it for a while, but series s? Heck no.

 

[Mozza]

The Switch 2 has been a massive success relative to expectations for its first year, but because the console was a known entity long before release, its hardware limitations were largely baked in. DLSS has certainly helped, but I am wondering if there is any untapped potential left to exploit going forward.

Frame Gen doesn't seem possible with the current hardware. While we will likely see incremental DLSS refinements, does anyone believe we can squeeze out a 30% performance improvement over the rest of the system's lifespan, or are we already at the pinnacle—meaning only a hardware revision could push things further?

Perhaps a Switch 2 OLED model would include a "Pro" spec bump, utilizing the new screen to finally hit those 4K/60 targets?

what's your take on the switch 2 moving forward!? Honestly interested!

It's going to be down to better optimization as developers have more time with the hardware.. I think hoping for a pro version is very optimistic.

 

[HerjansEagleFeeder]

My aunt that was my uncle last gen who works at Nintendo said there's a hidden RTX 3060 Ti in the S2's dock. You just wait for the official announcement of Majora's Mask Remake.

 

[Physiognomonics]

The very low bandwidth (compared to a PS4) will always be a big bottleneck in many games.

 

[Mr.Phoenix]

It's at worst 1.7 TF (less than PS4) and at best 3.6 TF (less than PS4 Pro), with 9GB of RAM, and at best ~100 GB/s bandwidth (also less than the PS4)... so no. It's not magic, and its saving grace is DLSS. That can only get you so far.

 

[hardwood007]

Of course, eventually they will get down to 6 or maybe 4nm. More per wafer = savings. If they can get the weight down further, that could save them some. It will lead to a more efficient console with maybe an extra hour of battery life.

The thing is, can they port it to a smaller node without redesigning the entire chip? The crux of it is that 8nm is a dead end node.

 

[PauloRoberto]

Without any real optimization by the devs? Absolutely not. Will SW2 sales (not hardware sales, game sales) be enough to justify this kind of commitment? I dont think so :-(

 

[Skifi28]

Wait until they activate the second hidden GPU on year 3.

 

[yogaflame]

Switch 2 might need a pro version 2 years from now

 

[Mozza]

My aunt that was my uncle last gen who works at Nintendo said there's a hidden RTX 3060 Ti in the S2's dock. You just wait for the official announcement of Majora's Mask Remake.

I was under the impression the Switch 2 dock has live hamster support.

 

[Mozza]

Switch 2 might need a pro version 2 years from now

I seriously doubt it will get it...

 

[deriks]

Frame gen is bullshit just to be clear

Still, devs have workarounds for the hardware. Have you see what some people are doing with the N64 and Dreamcast now?! It's fucking amazing. And it's not just by old hardware. Last of Us on PS3 is a looker, Ryse on Xbox One is a launch title that looks better than other big names NOW

Capcom adapted it's engine to run well on Switch 2, and if we take Code Veronica trailer in comparison with REmake 2, they found some dev shenanigans to get thing better, so it's expected to S2 also tske advantage of this

About a Pro version... Original GameBoy got it, DS got it, 3DS got it. Did You Know Gaming (can't remember the video) found that the riginal Switch was about to get it, but covid and more stuff turned into difficulty for this, so they released just the OLED version - that actually has a better hardware, but not enough to be called Pro

 

[SuperMac]

DLSS is where we will see performance upgrades. If Nvidia can update DLSS performance on the Switch 2. I thought I read some where, they said they could do this. Push those Tensor Cores baby!!!

 

[Madserb2023]

Doubt it. Consoles like the Switch 2 have hardware that is essentially a known quantity before they even come out. It's the reason why Rift Apart is still one of the best looking PS5 games. We still act like its the the early 2000s when you didn't even have documentation and it was all guess work.

 

[DarkShadowXI]

I seriously doubt it will get it...

Yeah there's no way

Switch 3 will come out in like 6-8 years. I don't see them straying away from Switch, though.

 

[KINGMOKU]

Switch 2 might need a pro version 2 years from now

Just like the original switch. I remember all the calls that support would die off, and sales would slow because it was so underpowered. Real gamers notice this kind of thing.

Glad they did release a switch pro, as now its guaranteed that the OG will become the best selling videogame system of all time.

Now we just sit back and wait for switch 2 pro.

 

[Fafalada]

uh... not really.
the times of devs learning hardware over years of optimising stuff for it are long over.

In fairness a lot of that has to do with 'optimising stuff' having moved away from hardware some time ago.
The scale of development is so far removed from hardware now, there's no space left for interesting optimizations (and markets/customers wholly embraced this - through Pro, incremental generations and PC segments alike).

Also this isn't about 'unknown' optimizations so much as it is 'time it takes to produce/maintain custom optimization paths'. If it costs me 12 months of development time that could be spent on the sequel/remake/spin-off/DLC instead of the 'main' release, and with no tangible impact on sales, why would I bother?
There's a sliver of chance AI could enable some of this in the future again (bringing the timescales down enough to make it viable) but again - market priorities seem to be wholly on the 'how do we do more of the same things we already do just more faster' and NOT 'how do we bring innovation in with all the efficiency gains' so - I'm not convinced.

We still act like its the the early 2000s when you didn't even have documentation and it was all guess work.

As I mention above - guess work isn't really the main issue - a lot of it is known quantities - time constraints are. A lot of 'possible' optimizations are left on the cutting room floor because noone has the time for it (and incentives just aren't there to 'must' have any of it).