• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

DLSS

DLSS stands for Deep Learning Super Sampling. The very quick and easy explanation of what it does is that it renders your game at a lower screen resolution, this gives you a performance boost as lower resolutions run faster. And then it applies a clever up-sample technique to turn your lower resolution image into a higher resolution image. You can think of this as a way of getting high resolution output such as 4k with better performance. The draw back being that up-sampling is not perfect, you get an approximation so it may not look as good.

The more detailed answer is that the up-sample technique is a new one based one machine learning, where a super computer is trained to up-sample games and make the output look as nice as possible. This up-sampling not only targets higher resolution but it has a type of Anti-Aliasing built in called Super Sampling. Anti Aliasing smooths images to removed jagged lines and Super Sampling Anti Aliasing is one specific type of Anti Aliasing that is baked into this up-sample process. So it's kind of a blend, it's up-sampling to get higher resolution but with AA pre-applied. The main reason you might want to run DLSS is if you have a more budget orientated card which cannot run native 4k and you want 4k. Or if you want to apply very performance expensive effects like Ray Tracing which might normally only run at 1080p but still run in high resolution like 4k.
 
It uses AI to upscale your game from a lower resolution to give you far better FPS and image quality which is on par with running native (sometimes even better) but with frame rates which are much higher.

I always use it, there is no reason not to. If you want raytracing with decent frame rates it's basically a must.
 
So if i am using 3 x 1440p monitors its probable not for that
Never used it on multi monitors, I have 2 but I only use one when gaming. But it should work just the same. For instance if you play on a 4k monitor the game may render at 1440p for instance and then the ai uses DLSS to produce a 4k image from this and then you get 4k but at almost double the frame rate of native 4k.
 
I believe the DLSS algorithm that the super computers train has to be targeted at specific game resolutions. I don't think things like super wide and multi-monitor resolutions are supported yet, and honestly given their uncommon use and the effort to train these deep learning algorithms I wouldn't hold my breath either.
 
Ultrawide 3440x1440 should work great with DLSS 2.0 games, you can confirm that with a quick Google.

If it's the same ratios as it uses with 1440p/4k then it should mean your system renders the game at 1720x720 or 2294x960 (fag packet maths) and delivers performance as such.
 
Just keep in mind that only a small percentage of games that come out have it. Great to have, but until they find a way for it to work on all games or at least until all future releases have it then it ain't something to get overly excited about imo.

Great news is Cyberpunk 2077 has it :)
 
Back
Top Bottom