In recent years, the world has increasingly introduced artificial intelligence into its products, which is gaining more popularity in different sectors. For example, the same Siri can easily tell you what the weather is like now. In Spotify, you can fill your music library with similar songs to your saved songs. And Amazon can determine which next product you buy long before you even think about it.
This kind of high technology has entered the realm of video games. Recently, information about a new version of video cards Nvidia appeared on the Internet. Today we will look at what is DLSS Nvidia, as well as talk about its most important details and capabilities.
What is Nvidia DLSS: capabilities of this video card
With the release of almost every new game or its version, the quality of their graphics only increases and improves. Along with this, the process of the operations performed becomes more complicated. This complexity of ray tracing is due to the fact that current games have a higher resolution than the usual 1080p. Accordingly, here it is necessary to speed up the calculations during the rendering periods.
DLSS or Deep Learning Super Sampling is one of the best technologies from Nvidia. DLSS meaning graphics card uses deep machine learning to boost frame rates in video games. This requires the use of complex graphical calculations. But thanks to DLSS, gamers don’t have to worry about stable fps performance if they increase the settings or the resolution.
Also, this technology can allow you to change the picture resolution from 1080p to 4K. In this case, the loss in quality will be minimal. This feature removes the need to load the game in 4K (and this can have a bad effect on your computer unless of course, you are a hacker with powerful hardware). Although the game will still show that you are using 1080p, the scaling will change and the quality of the game will be higher.
When testing TAA vs DLSS, the second option will undoubtedly win, since their technologies are slightly different and with this the final result with the quality of games.
DLSS release date isn’t known yet. But three models of these video cards will be presented in the fall, but possibly later.
What is Deep Learning Super Sampling?
Now let’s figure out what Deep Learning Super Sampling means. This deep learning involves algorithms that are based on the “brain” of neural networks. Compared to past algorithms that have been used to improve quality, DLSS graphics generate more data for the neural network, which improves performance and results accuracy. In addition, these same neural networks are able to improve their performance based on their own labors.
Does Deep Learning Super Sampling support less than 4K? Of course! The technology can improve not only 1080p, but 720p and even 540p. By improving image quality, productivity is also increased.
How exactly is the picture smoothed?
Nvidia was unable to provide us with a complete description of this process. But we can draw a conclusion based on the properties of deep learning. DLSS explained that the technology trains and trains the neural network using two sets of images. One set contains frames with anti-aliasing and others without. Based on this data, the neural network is able to create a model of how the image smoothing works and then creates a certain algorithm in order to independently perform these actions. Simply put, the video card learns to add/change on frames in such a way that the result is a picture of much better quality. Antialiasing isn’t even required here.
When rendering the frame geometry of an image in a game, the final appearance of the image determines the number of pixels. As an example, we will show how it works in a triangle to make it easier to understand how the technology works. First, we use a grid of squares (pixels) of 4×4 sampling. Here it is immediately clear that the picture won’t be of the best quality, since the squares are too large for small details. Accordingly, it is in those squares where there will be fewer details that quality will be removed.
If you increase this mesh to 8×8, the quality will be much higher, and almost the entire triangle will be painted over. This is how DLSS technology improves resolution and picture quality.
What is DLSS 2.0: main differences with version 1.0
The previous version 1.0 was trained separately for each new game. It is because of this that the whole process took quite a long time. Yes, he could also transform a picture in 4K resolution from 1080p, but, of course, the picture was not as high-quality as in 2.0. It also had other flaws that exacerbated the improved graphics in games.
The new generation of DLSS Nvidia explained that a more accurate algorithm is used here, in which there are no restrictions during training. Due to the fact that tensor kernels are used here, this is a significant improvement in image quality.
Thanks for reading the article! We hope you find this information interesting. Don’t forget to subscribe to be the first to know about new features and products.
- Why did you like the article?!
- Ooh, you didn't like it? But why?