sc94597 said:
The RTX 2050, in the video you shared, isn't doing 1080p upscaled to 1440p. It is doing 1080p DLSS Quality, which is internally 720p. Internally, Switch 2 is averaging about 75% of the pixels of 1080p (at times it hits the 1080p target, at other times it goes as low as 720p.) The Switch 2's Quality mode, at its worse, is comparable to the RTX 2050's image quality in the video you shared - roughly 44.4% of the pixels as native 1080p. Frame-rate is mainly a CPU-bound workload in a comparison like this. Switch 2's performance mode has shown us as much. So it's not the 2050 that is outclassing Switch 2 here, but whatever basic i5 is in the laptop. The texture difference is pretty obvious, where the textures are different. Digital Foundry showed as much in their video. Unless you think Indiana Jones is going to be running in the single digits, I really don't see what you're arguing here. Obviously the Switch 2 game is going to run better than the RTX 2050 can handle it, because the main issue is a lack of VRAM. |
How true is that though? Not questioning you, just always wanted to talk about this. Games require system ram and not just for OS.
The deck has like 16 gb ram, 4 for OS, 4 is vram other 8 is system.
A S2 has 12 gb and 3 is for the OS... but that remaining 9 gb isn't vram. Half or more will be used as system ram. So really the S2 has 4 dedicated vram, which is what the 2050 has.
Or am I thinking of this wrong? Point I'm getting is people to think the S2 has ram being used by OS and vram... this isn't true.
The S2 doesn't have 9 gb dedicated to vram. At most it would be 6 gb, at most. Yeah?
Last edited by Chrkeller - on 22 August 2025|
i7-13700k |
|
Vengeance 32 gb |
|
RTX 4090 Ventus 3x E OC |
Switch OLED







