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

Forums - Nintendo - Even without updates, many Switch 1 games see massive improvement on Switch 2

Before Switch 2 was launched, it was unknown whether its emulation of Switch 1 games would run them better than their native hardware.

Now that the system is out in the wild, it turns out that even without updates, many games that struggled on Switch 1 run much better on Switch 2, sticking to their maximum framerate and resolution target much more closely:

Last edited by curl-6 - on 06 June 2025

Around the Network

As expected and the way it should be... Especially in the modern era with uncapped framerates, dynamic resolution and more.

Artificially limiting old games is an old-school idea.



--::{PC Gaming Master Race}::--

Wouldn't say unknown, we've known about this since the announcement.

https://gamrconnect.vgchartz.com/post.php?id=9584973

Recently made this thread to cover this topic before launch:

https://gamrconnect.vgchartz.com/thread/249287/do-you-know-how-switch-2-backwards-compatibility-works/



@Twitter | Switch | Steam

You say tomato, I say tomato 

"¡Viva la Ñ!"

Dynamic resolutions not scaling higher though but frame rates maintaining their ideal frame rate i.e. 30fps or 60fps. The ARM A78 series can run A57 code directly without problems although maybe a few tweaks for compatibility but the T239 is not directly Switch 1 compatible for graphics in fact its very different architecture so I'm guessing these paid for updates have re-written optimised T239 graphics routines hence they are charged for. A modded Mariko Switch 1 with the CPU boosted by 100% and the graphics boosted by 50% also have much better frame rates but the dynamic scaling is more forgiving on accelerated Switch 1s but of course that is the same graphics hardware just more powerful. Makes me think the Switch 2 has ample CPU performance for this and runs the code more directly but graphically not so much. There is some debate on this forum regarding CPU performance. I think the Switch 2 is around 3-4x more powerful in CPU performance than the original Switch considering the low capacity battery and power hungry 10/8Nm fabrication process plus the chipset the T239 was developed from but others are claiming somewhere around 8-10x the CPU performance. However whatever it is there is ample CPU performance now to increase these frame rates. Modded Switches that run their CPUs at 2Ghz a 100% boost get huge increases in frame rates similar to this i.e. 2x more powerful CPU performance. You can still see on that video that some of the most demanding Switch 1 games that really struggle to get decent frame rates and have quite low frame rates don't quite pull off the ideal frame rate on Switch 2. The upgrade packs also benefit from DLSS so can render at a lower resolution to allow for DLSS upscaling but I'm only assuming they do that as not heard that confirmed yet but have seen a few comments in videos where it looked like it was upscaled but nothing confirmed by developers in this regard.



bonzobanana said:

Dynamic resolutions not scaling higher though but frame rates maintaining their ideal frame rate i.e. 30fps or 60fps. The ARM A78 series can run A57 code directly without problems although maybe a few tweaks for compatibility but the T239 is not directly Switch 1 compatible for graphics in fact its very different architecture so I'm guessing these paid for updates have re-written optimised T239 graphics routines hence they are charged for. A modded Mariko Switch 1 with the CPU boosted by 100% and the graphics boosted by 50% also have much better frame rates but the dynamic scaling is more forgiving on accelerated Switch 1s but of course that is the same graphics hardware just more powerful. Makes me think the Switch 2 has ample CPU performance for this and runs the code more directly but graphically not so much. There is some debate on this forum regarding CPU performance. I think the Switch 2 is around 3-4x more powerful in CPU performance than the original Switch considering the low capacity battery and power hungry 10/8Nm fabrication process plus the chipset the T239 was developed from but others are claiming somewhere around 8-10x the CPU performance. However whatever it is there is ample CPU performance now to increase these frame rates. Modded Switches that run their CPUs at 2Ghz a 100% boost get huge increases in frame rates similar to this i.e. 2x more powerful CPU performance. You can still see on that video that some of the most demanding Switch 1 games that really struggle to get decent frame rates and have quite low frame rates don't quite pull off the ideal frame rate on Switch 2. The upgrade packs also benefit from DLSS so can render at a lower resolution to allow for DLSS upscaling but I'm only assuming they do that as not heard that confirmed yet but have seen a few comments in videos where it looked like it was upscaled but nothing confirmed by developers in this regard.

Let me get this straight: you've been posting for weeks in every thread on the forum claiming that Nintendo didn't send units to reviewers because Switch games were running poorly, you see that they actually run really well, and you still stick to your delusions? Please, you've been lecturing us for weeks about topics you clearly don't understand, and it's obvious from your posts that your thoughts are a jumble of disconnected ideas, yet you keep telling everyone else how things are. Please, read, learn a bit, and then, if you want, you can give technical opinions.



Around the Network
bonzobanana said:

Dynamic resolutions not scaling higher though but frame rates maintaining their ideal frame rate i.e. 30fps or 60fps. The ARM A78 series can run A57 code directly without problems although maybe a few tweaks for compatibility but the T239 is not directly Switch 1 compatible for graphics in fact its very different architecture so I'm guessing these paid for updates have re-written optimised T239 graphics routines hence they are charged for. A modded Mariko Switch 1 with the CPU boosted by 100% and the graphics boosted by 50% also have much better frame rates but the dynamic scaling is more forgiving on accelerated Switch 1s but of course that is the same graphics hardware just more powerful. Makes me think the Switch 2 has ample CPU performance for this and runs the code more directly but graphically not so much. There is some debate on this forum regarding CPU performance. I think the Switch 2 is around 3-4x more powerful in CPU performance than the original Switch considering the low capacity battery and power hungry 10/8Nm fabrication process plus the chipset the T239 was developed from but others are claiming somewhere around 8-10x the CPU performance. However whatever it is there is ample CPU performance now to increase these frame rates. Modded Switches that run their CPUs at 2Ghz a 100% boost get huge increases in frame rates similar to this i.e. 2x more powerful CPU performance. You can still see on that video that some of the most demanding Switch 1 games that really struggle to get decent frame rates and have quite low frame rates don't quite pull off the ideal frame rate on Switch 2. The upgrade packs also benefit from DLSS so can render at a lower resolution to allow for DLSS upscaling but I'm only assuming they do that as not heard that confirmed yet but have seen a few comments in videos where it looked like it was upscaled but nothing confirmed by developers in this regard.

Dude, for God's sake, give a rest.

You've been spamming this bullshit of yours across dozens of threads for months now, and you've been on this absurd crusade of yours since like 2012 across multiple websites, I can go back to Digital Foundry articles from over a decade ago and still see your nonsense in the comments.

It's unhealthy to be this obsessed, knock it off.

And learn to use paragraphs while you're at it.



Furukawa: Who did this to me?!? /s

Seriously though, this is how it's supposed to be, more power, better performance out of the gate.



bonzobanana said:

There is some debate on this forum regarding CPU performance. 

No there isn't.

bonzobanana said:

I think the Switch 2 is around 3-4x more powerful in CPU performance than the original Switch considering the low capacity battery and power hungry 10/8Nm fabrication process plus the chipset the T239 was developed from but others are claiming somewhere around 8-10x the CPU performance. 

About 3x more performant, per clock.
Which is a good generational uplift.

But developers will be using the new SIMD instructions that have been added since the old A57 cores which came out 13 years ago which will give it a significant edge in games that leverage it.

Keep in mind that there is less work to be done on the Switch 2's CPU than the Switch 1, decompression for example has it's own fixed function block, that was a task that would have to be done on the Switch 1's CPU and could use an entire CPU core just by itself... The best way to increase CPU efficiency and performance  these days is often done by giving the CPU less work.

The issue here is that you are ONLY fixed on the raw numbers, the Switch 2 is far more efficient at every level with smarter design choices thanks to nVidia's ridiculous R&D budgets.

bonzobanana said:

You can still see on that video that some of the most demanding Switch 1 games that really struggle to get decent frame rates and have quite low frame rates don't quite pull off the ideal frame rate on Switch 2. The upgrade packs also benefit from DLSS so can render at a lower resolution to allow for DLSS upscaling but I'm only assuming they do that as not heard that confirmed yet but have seen a few comments in videos where it looked like it was upscaled but nothing confirmed by developers in this regard.

Who cares if it uses DLSS? Let's judge the output image which will vary from game to game and even frame to frame.

I am a gamer who prefers a raw output over upscaling, but if a 720P upscaled image to 1080P looks good, then it simply looks good.

You need to start looking at the bigger picture and stop nit-picking on the outliers.



--::{PC Gaming Master Race}::--

Kynes said:
bonzobanana said:

Dynamic resolutions not scaling higher though but frame rates maintaining their ideal frame rate i.e. 30fps or 60fps. The ARM A78 series can run A57 code directly without problems although maybe a few tweaks for compatibility but the T239 is not directly Switch 1 compatible for graphics in fact its very different architecture so I'm guessing these paid for updates have re-written optimised T239 graphics routines hence they are charged for. A modded Mariko Switch 1 with the CPU boosted by 100% and the graphics boosted by 50% also have much better frame rates but the dynamic scaling is more forgiving on accelerated Switch 1s but of course that is the same graphics hardware just more powerful. Makes me think the Switch 2 has ample CPU performance for this and runs the code more directly but graphically not so much. There is some debate on this forum regarding CPU performance. I think the Switch 2 is around 3-4x more powerful in CPU performance than the original Switch considering the low capacity battery and power hungry 10/8Nm fabrication process plus the chipset the T239 was developed from but others are claiming somewhere around 8-10x the CPU performance. However whatever it is there is ample CPU performance now to increase these frame rates. Modded Switches that run their CPUs at 2Ghz a 100% boost get huge increases in frame rates similar to this i.e. 2x more powerful CPU performance. You can still see on that video that some of the most demanding Switch 1 games that really struggle to get decent frame rates and have quite low frame rates don't quite pull off the ideal frame rate on Switch 2. The upgrade packs also benefit from DLSS so can render at a lower resolution to allow for DLSS upscaling but I'm only assuming they do that as not heard that confirmed yet but have seen a few comments in videos where it looked like it was upscaled but nothing confirmed by developers in this regard.

Let me get this straight: you've been posting for weeks in every thread on the forum claiming that Nintendo didn't send units to reviewers because Switch games were running poorly, you see that they actually run really well, and you still stick to your delusions? Please, you've been lecturing us for weeks about topics you clearly don't understand, and it's obvious from your posts that your thoughts are a jumble of disconnected ideas, yet you keep telling everyone else how things are. Please, read, learn a bit, and then, if you want, you can give technical opinions.

Are you retarded or something no one is lecturing anyone it is a discussion about Switch 1 compatibility. The previous discussion was about why Nintendo were not sending out review units and there was a big firmware update on day one. Nintendo themselves gave warnings about some Switch 1 games may not be fully compatible or run poorly but there were other possibilities for why review units were with-held and the unit was severely limited in operation until updated with hte day one patch. We don't know how important that day one firmware update was because we haven't seen the earlier firmware version. The hardware launched with a restrictive firmware.

It's pretty clear the video linked to stated good fame rates but no improvement in dynamic scaling. I personally thought it was interesting that frame rates were better but dynamic scaling was not improved as per the conclusions of that video. I can only guess you are nine years old or something to be so utterly childish in what you write. Just because you are a fanboy doesn't make your technical opinions superior, the Switch 2 is upscaling from very low resolutions to be competitive in performance and that is incredibly significant in any discussion about Switch 2 performance which many people seem to be ignoring because it doesn't suit their fanboy agenda. 

You don't just ignore what you don't like that isn't how discussing the specification of consoles works.



That's a great feature