By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Nintendo - How Will be Switch 2 Performance Wise?

 

Switch 2 is out! How you classify?

Terribly outdated! 3 4.55%
 
Outdated 1 1.52%
 
Slightly outdated 16 24.24%
 
On point 37 56.06%
 
High tech! 7 10.61%
 
A mixed bag 2 3.03%
 
Total:66
borisr on 06 February 2026
HoloDust said:
sc94597 said:

One of the reasons why I think there probably does need to be reworking (rather than just minor optimization) is the fact that an RX 6400 runs the game similar to a Series S despite being on paper somewhat weaker than a Series S. Basically runs the game at low settings with an internal 720p resolution at 30fps. So the PC version isn't that poorly optimized when theoretically worse hardware is achieving the same as the console version (which has additional optimizations.) 

An RX 6400 is about 90% of a Series S in theoretical performance. The Steam Deck is about 45% of an RX 6400 in terms of theoretical performance.

Ostensibly you could make that difference by scaling internal resolution to something like 45% of 720p, but we don't see that. Even at internally 260p the game isn't running at 30fps. Scaling isn't linear on the same architecture here.

So something in the render pipeline is bottlenecking performance on the Steam Deck. RT seems to be the natural answer of what that might be. The Steam Deck's CPU certainly isn't the issue because CPU utilization remains pretty low when playing the game and it has a decent CPU.

Maybe...but I guess we'll probably never know, since I doubt Ubi will decide to throw money down the drain and make custom Deck port.

It would be interesting to see what's the bottleneck, since, as I said, Ubi games tend to like AMD GPUs, apart from the fact that Deck, compared to Switch 2 in handheld mode, is not weak per se.

Deck is pc. You cant make custom port. You can make linux vulcan version and you can have same gpu optimisations as other pc games.



Around the Network
curl-6 on 07 February 2026

It is gonna be interesting to see how games like Rebirth and Oblivion hold up in their finished form.

Rebirth looked quite well converted in it's Direct showing, especially given it has another 4 months in the oven. Oblivion looked quite rough, but that's just "2026" at this stage so there's potentially still lots of time for it to be polished and optimized.

Now that I've seen it firsthand, Pragmata is another rather impressive title; granted it's more limited in scope, at least the demo is, but its one of the better looking and running current gen games on the platform thus far.

It is a bit disappointing though that we don't have any real exclusive showcase titles yet, with almost everything being a port from other consoles and Nintendo's own games playing it safe graphically.



bonzobanana on 08 February 2026
Pemalite said:
bonzobanana said:

I think part of the issue is the Switch 2 has no support chips like PS4 and Xbox One which are dedicated to processing network data. The PS4 has an ARM chip with 256MB of its own memory and works fully independently to give fantastic stable online play as does the Xbox One series with its south bridge chip.

Umm.

The Switch 2 uses a "SoC" known as a "System on a Chip".
That "System on a chip" is known as Tegra Orin.

Tegra Orin has a block on the SoC dedicated to handling networking operations known as the "MAC" or "Media Access Controller" for 10/100/1000 BASE-T Ethernet MAC capabilities.

It also has a block known as the "CAN FD Controller" for communication in automotive industries.

Here is the datasheet for the Tegra chip so you can update your understanding of what the SoC is actually capable of.
https://connecttech.com/ftp/pdf/jetson_orin_nx_datasheet.pdf

Rest assured all modern ARM SoC's offload network processing onto a dedicated processing block, no external chips required.
Doing network processing on CPU cores is a last century idea and drives up power consumption and latency, which is literally the opposite of the design goals that ARM SoC's strive for.

Sony was inefficient with the PS4, hence the external DRAM and ARM Cores, something the Xbox one didn't have, but this wasn't for network traffic itself, it was for managing background tasks and social functions.
Microsoft did it with less overhead on the Jaguar cores and system DRAM. - Both consoles had a south bridge to assist with managing I/O and networking as this is a holdover from PC designs.

bonzobanana said:
I'm certainly not saying anything controversial, Virtuos stated the CPU performance is at PS4 level (not pro) for the Switch 2. I'm actually stating going by the passmark score of 2000 that it is slightly above that as PS4 scores around 1700. Maybe with reduced cache it scores a bit less than 2000 because that benchmark was on 5Nm but I wouldn't of thought much less myself as they are only run at 1Ghz anyway and normally ARM A78s can go up to 3Ghz on 5Nm.

Keep in mind that the Playstation 4 also had 65% higher CPU clock than the Switch 2.

The Switch 2 cores themselves, clock for clock are absolutely superior to the Playstation 4... But we also need to remember that the Switch 2 SoC is doing more offloading of the CPU than the Playstation 4.

For example... Developers on the Playstation 4 would need to use the CPU in order to perform decompression tasks, the Switch 2 has a LZ4 decompression block on the SoC that takes that CPU load away, meaning more limited CPU cycles are available for actual gaming.

You need to start looking past the raw paper specs and start looking at the big picture, the Switch 2 SoC is capable of doing more, with less than the ancient Xbox One/Playstation 4 hardware... And the games are showcasing that, often returning better results... Imagine in another 7 years when developers have learned more about the hardware?

sc94597 said:

It's interesting to see Oblivion Remastered come to the Switch 2, in what seems to be a reasonable state, despite the SW2's very weak CPU. That was one game I was a bit concerned about not working well.

I also don't think the 8th Gen consoles could've ran it well. An FX 8300 (and most 4-core; 4-thread i5's) are a stutter-fest when playing this game. I have an i5 4690k + RTX A2000 itx build with DDR4 2400Mhz ram that I use for light image segmentation, and out of curiosity I installed the game on that and it basically stutters every 5-8 seconds after seeming like it would work pretty well at 30fps. Very much like when people would try to play the Witcher 3 or Dragon Age Inquisition with heavily overclocked Pentium G3258's that were popular in budget builds ten years ago. The i5 4690k is on paper +30% as performant as the SW2's CPU (more if we exclude the reserved cores) so that would've seemed like it wouldn't have worked well. 

This also puts Virtuos's comments in perspective. They probably had to do quite a bit of work to get this ported, given the Switch 2's CPU and that is why the developer honed in on it being the bottleneck, when asked. 

Edit: 

Also I am very impressed with FF7: Rebirth's trailer. It looks like it is consistent with Intergrade in terms of general end-result. 

Oblivion Remastered actually scales down on PC hardware super well.

It's the scripting and cell loading which causes the stutter and framerate issues, which is an I/O bottleneck and a characteristic of the Game Engine. (Gamebryo in this instance.)

I have zero concerns about the CPU/GPU being enough to run the game... Especially as I have a notebook with an RTX 2050 4GB that has returned good results. (Nothing compared to my desktop PC however.)

I am not sure how it runs on the Xbox Series X/Playstation 5 though as I haven't bothered to run the game on those consoles yet.

Tegra Orin is a SOC with up to 60W TDP meant for premium automobiles with a high cost and a huge feature set for the time. Yes it was designed around the same time as the Switch 2 chipset back in 2018/19 and shares the same fabrication node but its certainly not relevant to the Switch 2 spec with regard built in support chips which is a much more cut down mobile device designed for a TDP of less than 10W likely around 6W. I don't understand why you have mentioned this it adds nothing to the conversation and isn't relevant at all. Yes ARM A78s are typically much more powerful than PS4 Jaguar cores but not if you run them at 1Ghz they are designed to be fabricated on a 5Nm node ideally with a speed of up to 4Ghz (actively cooled). Frankly they are poor performing at 1Ghz and we know the cache will be reduced to be fabricated on a mainly 10Nm node. Anyway there are benchmarks that clearly indicate their performance and this backed up by Virtuos in that the overall speed is a bit faster than PS4 which is exactly where a passmark of 2000 sits in CPU performance. Surely with all the data out there and the developers themselves stating this its not really debatable. I've got a Razer Edge android games console that I picked up for £100 and that has 3 ARM A78 cores at 2.4Ghz on the optimal 5Nm node plus 4x A55 at 2Ghz and 1 super core X1 at 3Ghz which is competitive with high performance x86 CPUs. All those together works out at about a passmark of 4000 for CPU performance, still less than half that of the Steam Decks 9000 figure. The Switch 2 has very weak CPU performance of less than a quarter of steam deck. It's strengths are all in its GPU not CPU performance.

Orin

Nvidia announced the next-gen SoC codename Orin on March 27, 2018, at GPU Technology Conference 2018.[157]

It contains 17 billion transistors and 12 ARM Hercules cores and is capable of 200 INT8 TOPs @ 65W.[158]The Drive AGX Orin board system family was announced on December 18, 2019, at GTC China 2019.

Nvidia has sent papers to the press documenting that the known (from Xavier series) clock and voltage scaling on the semiconductors

and by pairing multiple such chips a wider range of application can be realized with the thus resulting board concepts.[159]In early 2021, Nvidia announced the Chinese vehicle company NIO will be using an Orin-based chip in their cars.[160]

The so far published specifications for Orin are:

  • CPU: 12× Arm Cortex-A78AE (Hercules) ARMv8.2-A (64-bit)[161][162]
  • GPU: Ampere-based, 2048[163] CUDA cores and 64 tensor cores1; "with up to 131 Sparse TOPs
    of INT8 Tensor compute, and up to 5.32 FP32 TFLOPs of CUDA compute."[164]
    • 5.3 CUDA TFLOPs (FP32)[165]
    • 10.6 CUDA TFLOPs (FP16)[165]
  • Samsung 8N process (derived from 8LPU)[165][166]
  • 275 TOPS (INT8) DL[165]
    • 170 TOPS DL (INT8) via the GPU
    • 105 TOPS DL (INT8) via the 2x NVDLA 2.0 units (DLA, Deep Learning Accelerator)
  • 85 TOPS DL (FP16)[165]
  • 5 TOPS in the PVA v2.0 unit (Programmable Vision Accelerator for Feature Tracking)
  • 1.85 GPix/s in the ISP unit (Image Signal Processor, with native full-range HDR and tile processing support)
  • Video processor for ? GPix/s encoding and ? GPix/s decode
  • 4× 10 Gbit/s Ethernet, 1× 1 Gbit/s Ethernet
1 Orin uses the double-rate tensor cores in the A100, not the standard tensor cores in consumer Ampere GPUs.

Nvidia announced the latest member of the family, "Orin Nano" in September 2022 at the GPU Technology Conference 2022.[167]

The Orin product line now features SoC and SoM (System-On-Module) based on the core Orin design and scaled for different uses from 60W all the way down to 5W. While less is known about the exact SoC's that are being manufactured, Nvidia has publicly shared detailed technical specifications about the entire Jetson Orin SoM product line. These module specifications illustrate how Orin scales providing insight into future devices that contain an Orin derived SoC.

https://en.wikipedia.org/wiki/Tegra



Pemalite on 08 February 2026
bonzobanana said:

Tegra Orin is a SOC with up to 60W TDP meant for premium automobiles with a high cost and a huge feature set for the time. Yes it was designed around the same time as the Switch 2 chipset back in 2018/19 and shares the same fabrication node but its certainly not relevant to the Switch 2 spec with regard built in support chips which is a much more cut down mobile device designed for a TDP of less than 10W likely around 6W. I don't understand why you have mentioned this it adds nothing to the conversation and isn't relevant at all. Yes ARM A78s are typically much more powerful than PS4 Jaguar cores but not if you run them at 1Ghz they are designed to be fabricated on a 5Nm node ideally with a speed of up to 4Ghz (actively cooled). Frankly they are poor performing at 1Ghz and we know the cache will be reduced to be fabricated on a mainly 10Nm node. Anyway there are benchmarks that clearly indicate their performance and this backed up by Virtuos in that the overall speed is a bit faster than PS4 which is exactly where a passmark of 2000 sits in CPU performance. Surely with all the data out there and the developers themselves stating this its not really debatable. I've got a Razer Edge android games console that I picked up for £100 and that has 3 ARM A78 cores at 2.4Ghz on the optimal 5Nm node plus 4x A55 at 2Ghz and 1 super core X1 at 3Ghz which is competitive with high performance x86 CPUs. All those together works out at about a passmark of 4000 for CPU performance, still less than half that of the Steam Decks 9000 figure. The Switch 2 has very weak CPU performance of less than a quarter of steam deck. It's strengths are all in its GPU not CPU performance.

Orin

Nvidia announced the next-gen SoC codename Orin on March 27, 2018, at GPU Technology Conference 2018.[157]

It contains 17 billion transistors and 12 ARM Hercules cores and is capable of 200 INT8 TOPs @ 65W.[158] The Drive AGX Orin board system family was announced on December 18, 2019, at GTC China 2019.

Nvidia has sent papers to the press documenting that the known (from Xavier series) clock and voltage scaling on the semiconductors

and by pairing multiple such chips a wider range of application can be realized with the thus resulting board concepts.[159] In early 2021, Nvidia announced the Chinese vehicle company NIO will be using an Orin-based chip in their cars.[160]

The so far published specifications for Orin are:

  • CPU: 12× Arm Cortex-A78AE (Hercules) ARMv8.2-A (64-bit)[161][162]
  • GPU: Ampere-based, 2048[163] CUDA cores and 64 tensor cores1; "with up to 131 Sparse TOPs
    of INT8 Tensor compute, and up to 5.32 FP32 TFLOPs of CUDA compute."[164]
    • 5.3 CUDA TFLOPs (FP32)[165]
    • 10.6 CUDA TFLOPs (FP16)[165]
  • Samsung 8N process (derived from 8LPU)[165][166]
  • 275 TOPS (INT8) DL[165]
    • 170 TOPS DL (INT8) via the GPU
    • 105 TOPS DL (INT8) via the 2x NVDLA 2.0 units (DLA, Deep Learning Accelerator)
  • 85 TOPS DL (FP16)[165]
  • 5 TOPS in the PVA v2.0 unit (Programmable Vision Accelerator for Feature Tracking)
  • 1.85 GPix/s in the ISP unit (Image Signal Processor, with native full-range HDR and tile processing support)
  • Video processor for ? GPix/s encoding and ? GPix/s decode
  • 4× 10 Gbit/s Ethernet, 1× 1 Gbit/s Ethernet
1 Orin uses the double-rate tensor cores in the A100, not the standard tensor cores in consumer Ampere GPUs.

Nvidia announced the latest member of the family, "Orin Nano" in September 2022 at the GPU Technology Conference 2022.[167]

The Orin product line now features SoC and SoM (System-On-Module) based on the core Orin design and scaled for different uses from 60W all the way down to 5W. While less is known about the exact SoC's that are being manufactured, Nvidia has publicly shared detailed technical specifications about the entire Jetson Orin SoM product line. These module specifications illustrate how Orin scales providing insight into future devices that contain an Orin derived SoC.

https://en.wikipedia.org/wiki/Tegra

Yes the Tegra in the Switch 2 is a deviation from the Tegra Orin, but it's not that much of a fucking deviation that was a chip built from the ground up which omitted fixed-function processing blocks, it's based on an existing design with a ratified stencil pattern, nVidia still needs to stick with it's establish design rules so they don't need to get the chip re-evaluated by outlets like FCC/IEEE/IEC/CISPR and more which takes time and money.
But this is all common sense and I am sure you were apprised of that, right?

Every ARM SoC for over a decade has literally included a network processing engine so as not to burden the CPU and save on power.
Same goes for Audio.
Same goes for video.
ARM SoC's were always designed around the concept of fixed function blocks to offload specialized tasks to save on power and increase performance.

If you disagree with that... Then I am sorry. But you are just wrong. - Your copied/pasted rubbish neither confirms nor refutes my previous post in any capacity.

As for the CPU performance... Again... If you even bothered to read my actual post, even if the Switch 2 CPU was equivalent to Jaguar CPU, the Switch 2 will be faster as it's offloading things like compression to a fixed function block.
As for the TDP argument, it's irrelevant. Chips operate on a voltage curve. You can have a single chip go from 10w to 1,000w if you want.

And again... You need to look beneath the black and white paper specifications of hardware and get a more nuanced, low-level understanding.




www.youtube.com/@Pemalite

curl-6 on 09 February 2026

DF have taken a look at FF7 Rebirth's showing and came away impressed; apparently it's a base 720p with DLSS, things like scene clutter, hair, and foliage are cut back from PS5, but overall they felt it holds up well and is notably better than the game's showing on Steam Deck. Framerate has some issues, but this was also true of FF7 Remake in pre-release footage and it was solved by release. 

Just going by eye, it also looks sharper than the PS5's performance mode, which was notably blurry.

Pragmata they are also positive about; pixel counts are lowered to hit 60fps, with DLSS upscaling from 540p in this case, but they describe the end result as "surprisingly good looking" and speculate that it retains RTGI, though RT reflections are cut and the hair is simplified.



Around the Network
Pemalite on 09 February 2026

I think UE5 is probably going to be the Switch 2's Achilles heel long term, especially games pushing Lumen and Nanite... But it's also a terrible engine on PS5 and Xbox Series X, so the fact the Switch 2 is presenting it so well is a testament.

Final Fantasy 7 Remake is definitely a step up over Steam Deck, but a step down from the Rog Ally X which can provide a proper 1080P output with higher settings... Which I think being a UE4 engine powered game tends to lend itself to Radeon more than Ampere thanks to it's more rasterization focused hardware.




www.youtube.com/@Pemalite

borisr on 11 February 2026
HoloDust said:
sc94597 said:

I already expessed what I think needs to be done. Ray-tracing needs to be replaced with SSR in a lot of places , and where it stays extant it needs to reduce in quality. Basically a combination of the Series S and SW2 modifications (Series S replaces with SSR or it is absent more often than SW2; SW2 reduces ray resolution and stability when extant. Steam Deck would do both.) I also don't think internal 540p, like SW2 handheld, is a reasonable goal unless FSR is discarded for some lighter AA - which probably would be for the best to be honest. If FSR is retained 400p would probably be a good target. 

I think where Steam Deck could be superior to SW2 is that it could retain mesh quality and asset density that have been reduced on SW2 to largely save on CPU resources. Alternatively or additionally, it can also potentially hit a 40fps mode easier, if the SW2 version is indeed CPU-bottlenecked. And of course there would be performance gains by cutting out the compatibility layer. 

All of this (along with the fact that the game has other issues, like SD controller not working) is probably why a year out there hasn't been performance updates for Steam Deck verification or to improve performance on other PC handhelds, unlike AC Shadows that released more recently. Heck AC Shadows even had a native Mac version and is coming to IpadOS

Well, we'll probably disagree on amount of RT reduction (I don't think that currently lighting is as bad on PC lowest settings as it is on Switch 2 with all the instability and "splotchiness", to use DF jargon, so Deck also having Switch 2 level of lighting would certainly help), but in principle yeah, that's what I'm talking about - custom port, as if Deck was actual console, with its strength and weaknesses, direct Vulkan code and all that jazz, instead of just handheld PC that's fed general PC port. Because, if you treat it just as a PC, it's more than natural that it can't keep up vs dedicated handheld that's in about the same performance rank.

Direct vulcan code is same as usual pc. No custom api.



Soundwave on 12 February 2026
Pemalite said:

I think UE5 is probably going to be the Switch 2's Achilles heel long term, especially games pushing Lumen and Nanite... But it's also a terrible engine on PS5 and Xbox Series X, so the fact the Switch 2 is presenting it so well is a testament.

Final Fantasy 7 Remake is definitely a step up over Steam Deck, but a step down from the Rog Ally X which can provide a proper 1080P output with higher settings... Which I think being a UE4 engine powered game tends to lend itself to Radeon more than Ampere thanks to it's more rasterization focused hardware.

ROG Ally X again is double the cost and far bulkier of a device, one would hope it can run something better, if a console was double the cost of a PS5 it would be a flat out embarrassment if it couldn't run games notably better. If I pay double for a GPU, I obviously also would expect better performance. 

Last edited by Soundwave - on 12 February 2026

Pemalite on 12 February 2026
Soundwave said:
Pemalite said:

I think UE5 is probably going to be the Switch 2's Achilles heel long term, especially games pushing Lumen and Nanite... But it's also a terrible engine on PS5 and Xbox Series X, so the fact the Switch 2 is presenting it so well is a testament.

Final Fantasy 7 Remake is definitely a step up over Steam Deck, but a step down from the Rog Ally X which can provide a proper 1080P output with higher settings... Which I think being a UE4 engine powered game tends to lend itself to Radeon more than Ampere thanks to it's more rasterization focused hardware.

ROG Ally X again is double the cost and far bulkier of a device, one would hope it can run something better, if a console was double the cost of a PS5 it would be a flat out embarrassment if it couldn't run games notably better. If I pay double for a GPU, I obviously also would expect better performance. 

The Switch 2 is double the price and is bulkier than the Switch Lite.
We can move that goal post around all day.

You want the better device with better hardware? You ironically need to pay for it, who would have thought?




www.youtube.com/@Pemalite

Soundwave on 12 February 2026
Pemalite said:
Soundwave said:

ROG Ally X again is double the cost and far bulkier of a device, one would hope it can run something better, if a console was double the cost of a PS5 it would be a flat out embarrassment if it couldn't run games notably better. If I pay double for a GPU, I obviously also would expect better performance. 

The Switch 2 is double the price and is bulkier than the Switch Lite.
We can move that goal post around all day.

You want the better device with better hardware? You ironically need to pay for it, who would have thought?

And the Switch 2 is a waaaaaaaaaay better device than the Switch 1 Lite, a full generation ahead easily of the Switch Lite, the ROG Ally X can run games moderately better than the Switch 2 but is no where near a generational uplift while needing to be much bulkier and having a smaller screen for more than double the cost. No detachable Joycons either. 

At $1000 it's a niche, niche device, if that's where you have to go to get better than Switch 2 performance in a still bulkier form factor, than Nintendo delivered good hardware bang for their buck. 

The Switch 1 Lite would be more like a Playstation 1 in your analogy, the Switch 2 being a PS2, and the ROG Ally X being an original XBox (2001), but it would be if the XBox cost $1000 while the PS2 was $450. 

Last edited by Soundwave - on 12 February 2026