DLSS | How to get extra Performance From Your Games | NVIDIA

You might’ve of heard of  NVIDIA DLSS, but do you know what it is? Well, look no further because this blog will look at the three current versions of DLSS, what they do and which games currently support them. So sit back, relax and let’s get you some more precious frames in your games.

What is NVIDIA Deep Learning Super Sampling?

Deep learning super sampling (DLSS) is a video rendering technique developed by NVIDIA, that aims to boost performance and frame rates by rendering frames at a lower resolution than displayed on your monitor. Then, using deep learning—a form of AI— upscales the frames so they look as clear and sharp as intended at the native resolution. Meaning higher frame rates are more achievable. For example, with DLSS, a game’s frames could be rendered at 1080p, then upscaled and output at 1440p, giving the user the higher frame rates like they could achieve at 1080p whilst keeping the graphics quality close to 1440p.
There are alternative rending methods, such as temporal anti-aliasing (TAA), which is a post-processing algorithm. To put this into layman’s terms, DLSS is just a more advanced TAA.

At the time of writing this blog, there are currently three versions of DLSS:
• DLSS
• DLSS 2
• DLSS 3
But what is the difference between the three? What hardware do you need to take advantage of this wonderful bit of AI-based technology?

Version 1

DLSS is the first iteration of the technology, as mentioned before, a successor to TAA. NVIDIA claimed that DLSS’ temporal stability and image clarity was superior to the commonly used TAA. When paired with deep learning AI, this groundbreaking tech was the baseline for what was to come. Being the first generation of this technology, version 1 has had to train on a game-to-game basis, making it the least supported in games.

Version 2

The first significant improvement of the second generation was that DLSS 2.0 didn’t have to be trained on a game-by-game basis. Instead, it uses a single, more extensive, nongame-specific network. This was great news to all gamers. The previous point meant DLSS 2.0 could be implemented into games much more effortlessly, and developers could take full advantage of this fantastic technology.

The next improvement for this generation was the ability to control which mode DLSS is set to whilst gaming. Unlike the original version, which only had an on-and-off option. DLSS 2.0 has three options to choose from, quality, balanced and performance, but what is the difference?

Performance– up to 70%+ higher FPS, but slightly worse perceptible quality than native resolution.

Balanced – up to 40% higher FPS, and basically looks like native resolution. Some things can be visibly worse, but others are better (graphics can look sharper).

Quality – up to 25% higher FPS, and some things are evidently better than native. Almost nothing is worse than native resolution.

Again these are rough numbers and depend on several factors such as the game, resolution, and if you have a CPU bottleneck.

The improved Tensor core usage optimisation is the final improvement for this generation of DLSS. NVIDIA stated, “A new AI network more efficiently uses Tensor Cores to execute 2X faster than the original, ” meaning it takes half the time to reconstruct details in the image.

Want to take advantage of DLSS Version 1 & 2 Technology?

Version 3

DLSS 3 is this revolutionary tech’s latest and greatest instalment, boasting new features.

The most groundbreaking addition to DLSS 3 is the Optical Multi Frame Generation, to generate entirely new frames, but what does this actually mean? Previously, DLSS would generate new pixels to achieve higher FPS. NVIDIA says, “Optical Multi Frame Generation generates entirely new frames, rather than just pixels, delivering astounding performance boosts. The new Optical Flow Accelerator incorporated into the NVIDIA Ada Lovelace architecture analyses two sequential in-game images. Then calculates motion vector data for objects and elements that appear in the frame but are not modelled by traditional game engine motion vectors. This dramatically reduces visual anomalies when AI renders elements such as particles, reflections, shadows and lighting.” In basic terms, NVIDIA’s latest GPUs calculate the motion of objects that appear in the frames. Then generates new frames based on this giving the user performance like never seen before whilst using this technology.

The next improvement is the incorporation of NVIDIA Reflex, which “synchronises the GPU and CPU, ensuring optimum responsiveness and low system latency.” The lower the latency, the more responsive the game’s controls are, giving you an edge over the competition (who aren’t using DLSS 3 and NVIDIA Reflex).

There are a couple of downsides to DLSS 3, the first being the lack of games. Due to how new this version of DLSS is, only 35 games (at the time of writing this) support the latest version. The second downside is that it is only available on the new NVIDIA GeForce RTX 40 Series, unlike DLSS 2, which was compatible with any RTX GPU. This is due to DLSS 3 being reliant on the latest NVIDIA Ada Lovelace architecture and currently is not backwards compatible with previous architectures.

Want to take advantage of DLSS 3 Technology?

What games support DLSS?

Now you know what DLSS is, the real question is how you utilise it and which games utilise DLSS? Well, we have you covered. Below is every game to date that allows you to boost your performance using NVIDIAs groundbreaking deep learning AI technology.

Recommended Games for DLSS 1 & 2

Recommended Games using DLSS 3

As you can see, hundreds of games support DLSS, meaning you’ll find it hard not to get extra performance from your favourite games.

How do I Turn on DLSS?

Now we know what DLSS is and which games support this excellent tech. The real question now is how do you utilise it? Every game’s options are slightly different, but here’s a basic overview of where you must look to enable DLSS

1. Load one of the games on the list above.
2. Go to your graphics settings and look for NVIDIA DLSS
3. Turn on DLSS
4. Choose which mode you want (quality, balanced or performance)
5. Apply the settings
It’s that easy. Now it’s time to enjoy all those extra frames in your favourite games.

Overall, NVIDIA DLSS is the best way to get extra performance from your system whilst gaming, making games like Cyberpunk 2077 run at higher frames without sacrificing graphics quality. Don’t worry if you don’t have an NVIDIA RTX GPU. You can use our system configurator to find the perfect PC for your gaming needs and use your new RTX GPU to benefit from DLSS in all your favourite games.