Digital Foundry: Cyberpunk 2077 Console Benchmarks: Switch 2/ Xbox Series S/ PS4/ PS4 Pro!

adamsapple

Or is it just one of Phil's balls in my throat?



We've previously used CD Projekt RED's internal benchmarks to stack up Switch 2 against Steam Deck, but what about PlayStation 4, PS4 Pro and - of course - Xbox Series S. Comments from Virtuos satellite studio Black Shamrock suggest that Switch 2 features GPU power comparable with Series S and CPU power just a touch higher than PlayStation 4 - but to what extent does that tally with Cyberpunk 2077's outputs? It's just one game, running with different settings on each system, but nevertheless, the results are intriguing.


00:00 Previously on Digital Foundry
00:19 Introduction - The Black Shamrock Interview
03:04 Cyberpunk 2077 Test Sequences
03:42 Benchmarks - GPU Performance Focus
13:08 Benchmarks - CPU Performance Focus
19:00 And Now, The Conclusion
 
- NS2 vs PS4 vs PS4 Pro vs Series S comparison
- Testing done using a benchmark module in the game that can be replicated identically

- GPU focused testing

- Testing 1 - 30 FPS testing:
- PS4 dropped 4 frames, PS4 Pro 2, NS2 and Series S stick to 30fps target entirely
- The main difference lies within DRS range.
- Sample 1 DRS: PS4: 864p, Series S, NS2, PS4 Pro: 1080p with DRS possibility on Series S to 1440p
- Sample 2 DRS: PS4: 864p, Pro: 1188p, NSw: 945p, Series S: 1440p

- Testing 2 - 60FPS testing (more like 120hz)
- NS2 running at 40fps 660p, Series S running at 60fps is 792p
- PS4 and Pro do not support 120hz (or 60fps regardless)
- NS2's DLSS computational load is higher than what PS4 or Pro can do

- Kabuki streaming test for CPU/GPU strain:-

- 30fps testing:
- Starting areas, all consoles run the 30fps mode practically locked, NS2 sees minor drops to 28fps
- NS2, however, streams in a lot faster and farther than PS4/Pro
- End of the run denser area sees PS4 drop to mid 20's while Pro stays at 30
- DRS sees NS2 and Series S at 1080p and PS4 and Pro around 900p
- Series S seems to stick to 1440p @ 30fps in the lighter areas.
- Later denser areas show Series S at 720p/60fps with NS2 being 790p but around high 20's

- 120hz testing:
- 120hz mode in above testing shows NS2 DRS 810p while at 40fps
- Series S is 900p at 60fps in same 120hz testing
- Later areas in the Kabuki testing shows a more dynamic DRS range on both

- Overall conclusion:
- NS2's docked performance over PS4 shows average 32% resolution improvement
- Similar resolution to PS4 Pro but NS2 is using DLSS
- Series S in quality mode can see as high as 70% resolution increase and up to 73% in performance mode while targeting higher frame rate (60 vs 40)

---


CPU specific testing:

- NS2 saw similar drops like the Steam Deck when CPU stressed
- PS4 Pro and Series S have no problem keeping 30fps maintained where NS2 and PS4 can buckle
- In Performance Mode, Series S can double performance over NS2 in stress areas (28 vs 60fps)
- NS2 has an advantage in streaming and draw distances over PS4/Pro, however

Conclusion:
- DF thinks Virtuos's comments about games being easy to port to NS2 aren't exactly 'born out' in their testing, they conclude that the games may be easily portable but not anywhere near the same performance.
 
Last edited:
Bout how I figured, also the PS4 Pro likely is not even optimized and I am sure could do better.

Series S clearly the winner, though no one is shocked. Nice handheld port for NS2 though.
 
Last edited:
Isn't this like a repeat of all other tests?

Edit: Nvm, it isn't, but they really like talking about this game
 
Last edited:
I've been playing the switch 2 version and the PS5 Pro version simultaneously and toggling back and forth on my HDMI input and tbh the DLSS on the switch 2 makes the image look more pleasing at least when still, compared to the PS5 Pro, it looks less oversharpened if that makes sense.
 
I've been playing the switch 2 version and the PS5 Pro version simultaneously and toggling back and forth on my HDMI input and tbh the DLSS on the switch 2 makes the image look more pleasing at least when still, compared to the PS5 Pro, it looks less oversharpened if that makes sense.
That's what I'm thinking too, its beating PS5 PRO in many ways, let alone the consoles this video is comparing them with.
 
As Rich said, the dev team behind that interview don't know shit. CPU between PS4 and PRO, GPU not even near to Series S, while DLSS doing the heavy lift.
 
They finally tested the Pro.
I've been playing the switch 2 version and the PS5 Pro version simultaneously and toggling back and forth on my HDMI input and tbh the DLSS on the switch 2 makes the image look more pleasing at least when still, compared to the PS5 Pro, it looks less oversharpened if that makes sense.
That's what I'm thinking too, its beating PS5 PRO in many ways, let alone the consoles this video is comparing them with.
There is no PS5 Pro version. Isn't it 4k @ 30fps on PS5?
Edit: It's 1440p? No, Switch 2 is not going to look better. The 60fps mode is a higher than 1440p.
 
Last edited:
I hope the results can now put these useless discussions to bed. The Switch 2 is absolutely no where close to the series is. It's much closer to the ps4 than it is to the series s. In fact, the switch 2 is not even better than the ps4 pro.

None of this surprising as the Switch 2 is a handheld.
 
They conclude the console is between base PS4 and PS4 Pro.

I mean Switch docked is 3.1 Tflop, the PS4 is 1.5Tflops and PS4 Pro is 4Tflops, so obviously it's between the two.

The most interesting part is that DLSS' cost have an impact on performance.
 
Conclusion:
- DF thinks Virtuos's comments about games being easy to port to NS2 aren't exactly 'born out' in their testing, they conclude that the games may be easily portable but not anywhere near the same performance.

What? A company that gets paid to port games to NS2 exaggerated how easy it is to port games to NS2?

Oh My God Wow GIF by The Roku Channel
 
Did you forget the whole year CD Projekt had to optimize PS4 and PS4 Pro version to have the game back on PSN ? And that they kept updating it until like 2023 ?
Speaking on the Pro version specifically

But not sure, maybe they did.

The last gen consoles never got past 1.6 updates

2.0 was the game changer one with all the changes and far more fixes and such and the NS2 benefits from it
 
Lags behind XSS with better visuals?
04qgPlwiuV9pGjUW.gif
Just kinda going by this really

0fps testing:
- Starting areas, all consoles run the 30fps mode practically locked, NS2 sees minor drops to 28fps
- NS2, however, streams in a lot faster and farther than PS4/Pro
- End of the run denser area sees PS4 drop to mid 20's while Pro stays at 30
- DRS sees NS2 and Series S at 1080p and PS4 and Pro around 900p
- Series S seems to stick to 1440p @ 30fps in the lighter areas.
- Later denser areas show Series S at 720p/60fps with NS2 being 790p but around high 20's


- 120hz testing:
- 120hz mode in above testing shows NS2 DRS 810p while at 40fps
- Series S is 900p at 60fps in same 120hz testing
- Later areas in the Kabuki testing shows a more dynamic DRS range on both

- Overall conclusion:
- NS2's docked performance over PS4 shows average 32% resolution improvement
- Similar resolution to PS4 Pro but NS2 is using DLSS
- Series S in quality mode can see as high as 70% resolution increase and up to 73% in performance mode while targeting higher frame rate (60 vs 40)

But no, this game does NOT look better on any planet compared to a PS5 Pro

Not even on Uranus does it look better than a PS5 Pro, objectively.

Performance is LIGHT YEARS ahead of NS2 on XSS.... light years.
 
Last edited:
Just kinda going by this really

0fps testing:
- Starting areas, all consoles run the 30fps mode practically locked, NS2 sees minor drops to 28fps
- NS2, however, streams in a lot faster and farther than PS4/Pro
- End of the run denser area sees PS4 drop to mid 20's while Pro stays at 30
- DRS sees NS2 and Series S at 1080p and PS4 and Pro around 900p
- Series S seems to stick to 1440p @ 30fps in the lighter areas.
- Later denser areas show Series S at 720p/60fps with NS2 being 790p but around high 20's


- 120hz testing:
- 120hz mode in above testing shows NS2 DRS 810p while at 40fps
- Series S is 900p at 60fps in same 120hz testing
- Later areas in the Kabuki testing shows a more dynamic DRS range on both

- Overall conclusion:
- NS2's docked performance over PS4 shows average 32% resolution improvement
- Similar resolution to PS4 Pro but NS2 is using DLSS
- Series S in quality mode can see as high as 70% resolution increase and up to 73% in performance mode while targeting higher frame rate (60 vs 40)

But no, this game does NOT look better on any planet compared to a PS5 Pro

Not even on Uranus does it look better than a PS5 Pro, objectively.

Performance is LIGHT YEARS ahead of NS2 on XSS.... light years.
It looks better than XSS and comparable to PS5 Pro.
 
I'm thinking CDPR might actually be using x86 to ARM emulation here, the CPU can't possibly be this bad even at 1 GHz.
XSS cpu has higher powerdraw than whole docked switch2, bro, and 3,5x higher frequency on top of being 8cores/16threads, switch2 cpu is only 8cores total(6 avaiable for games, 2 for running operating system).
 
Effort wise would not be worth it, but power wise, it can easily do 2.0+phantom liberty, especially knowing they run on NS2/Deck etc
last gen consoles lack proper data streaming for that, not to mention that AFAIK (haven't played CP2077) 2.0 also brought higher graphical settings, and if it doesn't Switch 2 has higher graphical settings than even PS4 Pro
 
Yup. Worse shadow cascade (or whatever its called), jaggies that could saw your finger off.

I don't have a Series S, but I do have a PS5. Boot up the game on Switch 2 and take some pics.
The Series S alone takes it out back and spanks it

Whips it and smacks it around

The NS2 is a powerful handheld but on no planet does it trade blows with current gen. None. Why people thinks this is beyond me.

DF legit just showed the XSS 70% higher performance and DOUBLE the fps
 
Top Bottom