Tech Guided is supported by readers. If you buy products from links on our site, we may earn a commission. This won't change how much you pay for the products and it doesn't influence our decision in which products we recommend. Learn more

What is NVIDIA’s DLDSR? (And How to Enable It)

What is NVIDIA’s DLDSR And How to Enable It

While the Graphics Card shortage is speculated to be nearing its end, the future remains uncertain.

And even with NVIDIA’s 40xx series and AMD’s 6×50 cards just over the horizon, there’s a good chance these cards will be both expensive, and difficult to obtain.

With that in mind, many users have been looking for ways to squeeze as much performance as possible out of their current GPUs.

For owners of NVIDIA cards, DLDSR is one such feature that allows for improved performance at no extra cost. So what is it, and how do you enable it?

DLDSR vs DLSS

Deep Learning Dynamic Super Resolution (abbreviated as DLDSR) is similar to DLSS in the sense that it uses excess GPU power to improve both image quality and framerate. While there is a distinct difference between the two—DLDSR being a downscaling technique, and DLSS being an upscaling technique—they both make use of NVIDIA’s powerful AI, and by extension, the onboard tensor cores.

And while these are both powerful tools in their own right, tensor cores are only present in NVIDIA’s RTX series of cards—meaning that at the time of writing, DLDSR is limited to their 20 and 30 series of cards (though this will soon include the upcoming 40 series).

To understand how DLDSR works, it’s useful to also understand Dynamic Super Resolution (DSR), as well as Dynamic Super Sampling (DLSS).

What is DSR?

Of these three technologies, DSR is the oldest of the bunch, having been introduced seven years ago, along with the advent of NVIDIA’s 970 and 980 graphics cards. While the terms “upscaling” and “downscaling” might cause confusion, the process and results are rather simple.

GeForce Tech Demo: DSR

DSR essentially takes advantage of any available GPU headroom, allowing frames to be rendered at a higher resolution. These higher resolution images are then made smaller to fit the screen. In spite of this process, the improved picture quality resulting from the higher initial rendering resolution is noticeable.

This is similar to the way that Super Sample Anti-Aliasing (SSAA) works—though SSAA is something that must be implemented by game developers, rather than running off of the GPU itself.

What is DLDSR?

DLDSR serves as a direct upgrade to DSR itself, making use of NVIDIA’s neural network to provide the same end product as DSR while making the process more efficient, making it less taxing on the GPU overall. This is achieved through use of the onboard neural network of RTX cards.

DSR archives its end result by forcing the GPU itself to put in extra work, and in this process, much of this extra work ends up being thrown away. This is where the neural network and tensor cores of RTX cards step in. Rather than straight up rendering images at a higher resolution, the AI of RTX cards is able to understand what the image should look like at a higher resolution, and modifies the image accordingly before being displayed.

DLDSR vs DSR vs Native

It’s worth noting that while DLDSR provides increased frame rates in comparison to DSR, you’ll still see decreased frame rates the higher you set DLDSR.

What is DLSS?

As previously mentioned, unlike DSR and DLDSR, DLSS is an upscaling technique, rather than a downscaling technique. Instead of rendering the initial image at a higher resolution, DLSS renders it at a lower resolution, then makes use of the AI to infer what the image would look like at a higher resolution.

The results themselves can be a bit of a mixed bag, but unlike DSR and DLDSR, DLSS isn’t meant to be a standalone technique to improve performance. Instead, it’s meant to be used in tandem with raytracing, minimizing its performance reducing effects.

Also unlike DSR and DLDSR, DLSS requires support from the game itself, greatly limiting its applicability.

How Do You Enable DLDSR?

With all that out of the way, how do you enable DLDSR? Fortunately, it’s pretty simple.

Before anything else, make sure your drivers are up to date—DLDSR was only recently added, and keeping them up to date minimizes the risk of anything going wrong.

  1. First, launch the NVIDIA control panel. You can do this by right-clicking an empty space on your desktop area.
  2. Next, navigate to 3d settings.
  3. Under 3d settings, there should be a dropdown labeled “DSR – Factors.” Click on it.
  4. Select the DL-Based DSR factor of your choice.
  5. Click “OK,” and then “Apply.”

nvidia-control-panel-dldsr-configuration

This will add new resolutions to the NVIDIA control panel, Windows Display settings, and in-game resolution settings.

When Should You Use DLDSR?

While DLDSR (and by extension DSR) offers substantial benefits—both in terms of image quality and performance—it’s not something that you can or should be using at all times.

This is because, as previously mentioned, DLDSR makes use of GPU headroom, or extra GPU power that’s not actively being utilized. Because of this, if your GPU usage is at or close to 100% (think 90-100%) on certain games, it’s best to leave it off.

Conversely, if GPU usage is at or below ~80%, DLDSR just might help you get that extra bit of performance left on the table.

Robert Brandon

Robert has been building and selling computers as a hobby for a little over 3 years now. When he's not busy immersing himself in his studies, he spends his time reading, writing, and duking it out with others in a wide variety of multiplayer games.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.