What is NVIDIA DLAA and When Should You Use It?

A Geforce RTX 3050 Ti GPU inside a computer tower
Elizabeth Henges / How-To Geek

NVIDIA’s DLAA (Deep Learning Anti-Aliasing) is an AI-powered tool to eliminate jagged edges in video games, maintaining native resolution unlike DLSS. It improves image quality with less performance cost than traditional methods, ideal for GPUs with sufficient power to run a game at targeted resolution and framerate.

We live in a time where graphics cards are not just about brute rendering power. DLAA is a feature on modern NVIDIA cards that can increase image quality, but what exactly is DLAA, and when should you use it?

What Is DLAA?

In the simplest terms, DLAA (Deep Learning Anti-Aliasing) is a technique that uses AI to eliminate jaggies or the ‘staircase effect’ on the edges of objects in video games, a common problem that anti-aliasing techniques aim to solve.

The ‘deep learning’ part refers to the way that the system is trained. NVIDIA uses machine learning models that are taught how to apply anti-aliasing effectively using a vast library of high-definition images.

RELATED: What Is NVIDIA Image Scaling, and How Do You Use It?

The fascinating part of DLAA is that, unlike Deep Learning Super-Sampling (DLSS), it doesn’t involve any upscaling. While DLSS works by rendering games at a lower resolution and then using AI to upscale the image intelligently, DLAA operates on the game’s native resolution, focusing solely on anti-aliasing. In essence, DLAA brings the sharpness and clarity of DLSS, but without changing the resolution.

The Benefits of DLAA

DLAA can produce excellent anti-aliasing results while preserving more of the original image detail than traditional methods and NVIDIA’s own DLSS. Because the system is trained using machine learning, it can apply a sophisticated understanding of how real-world objects look to the aliasing problem.

RELATED: What is NVIDIA DLSS, and How Will It Make Ray Tracing Faster?

DLAA can potentially improve visual fidelity at lower performance costs compared to conventional anti-aliasing techniques. It uses the same dedicated machine learning hardware DLSS upscaling uses to accelerate its treatment of the image. So while DLAA still comes with a performance hit, it’s nowhere near as heavy as a more traditional method would have to be to achieve similar results.

When to Use DLAA Instead of DLSS

Diablo IV DLAA Toggle
In Diablo IV, you’ll find DLAA in the same options list as DLSS.

DLAA is not meant to replace DLSS; instead, it’s an additional tool in your arsenal. The choice between DLSS and DLAA will depend on your hardware configuration and the specific system requirements of the game you’re playing.

RELATED: Does It Matter Which Graphics Card Manufacturer You Choose?

You might choose DLAA over DLSS when your GPU has ample power to run a game at your target resolution and framerate, but you’d like to enhance the image quality by minimizing aliasing artifacts. In this scenario, DLAA can be a fantastic tool to maintain your current resolution and improve overall image quality.

On the other hand, if you’re struggling to hit your target resolution and framerate, then DLSS might be the better choice. DLSS can help your system punch above its weight by rendering the game at a lower resolution and then using AI to upscale the image, offering better performance.

It’s worth pointing out that DLSS is also effectively a form of anti-aliasing. So the choice between DLAA and DLSS isn’t really about aliasing at all, in my opinion. Instead, it elevates the image quality of the natively-rendered scene at a small performance cost. DLSS on the other hand, tries (and usually succeeds) at matching the quality of the native image plus traditional aliasing. However, with DLSS you get a welcome frame rate boost as well.

To give you a practical example, my gaming laptop can run Diablo IV at 1440p ultra preset using DLSS in quality mode, at frame rates in excess of 120fps. However, I personally don’t see the benefit of going above 60fps in an isometric action RPG, so instead I capped the frame rate at 60 and activated DLAA to maximize image quality with a smooth, locked frame rate.

Here is the game using DLSS at quality mode, targeting 1440p. I’ve used the game’s built-in zoom function to highlight finer details.

Diablo IV Ultra DLSS Quality 1440p

Here are the exact same scene and settings, except this time, we’re using DLAA. Can you tell the difference?

Diablo IV Ultra DLAA Quality 1440p

Apart from the FPS count in the corner there, it may not seem like there’s any difference at all. So let’s zoom into those rock formations at the top left corner, starting with the DLSS Quality mode.

Diablo IV Ultra DLSS Quality 1440p Zoomed

And now, here we have DLAA, with all other settings the same.

Diablo IV Ultra DLAA Quality 1440p Zoomed

You’d be forgiven for not being able to tell what DLAA is even doing, but of course screenshots don’t tell the whole story, since the two methods may perform differently in motion. DLSS may be more prone to motion artifacts, but whether you can actually see these during normal gameplay depends on you and your setup.

Different titles that support DLSS and DLAA may also have more variance between the image quality each technology can achieve in stills and motion. So don’t be afraid to flip between them and decide for yourself which offers the best tradeoff between quality and performance.

Our sample game here, Diablo IV, definitely looks sharper in motion with DLAA on, and given how much performance overhead I have above 60 FPS, the sharp stability in motion is worth the FPS tradeoff. However, in games such as competitive first-person shooters and racing titles, you’d probably prefer snappier responsiveness over slightly sharper image detail nine times out of ten. Whatever you choose, though, now you know the difference between DLSS

Leave a Comment