In the graphics card market, at the moment, two giant companies are fighting each other. On the other hand, we have NVIDIA, a green faction full of innovative ideas that have been available for years alongside computers and Intel processors. On the other hand, AMD bravely fights and grows – the red stronghold of the rebels, changing the gray-brown reality. Competing with each other, both companies have their own technologies, teams of engineers, and ideas for developing graphics display technologies. This is what we will cover in this article. I’d like to introduce you to what NVIDIA DLSS and AMD FidelityFX Super Resolution (FSR for short) really are and how they will dramatically change our perception of native resolution.
Let’s fight for teraflops, because thanks to them we will get original 4K
Even before the premiere of the consoles of the previous generation, we had a very intense fight between fans of Sony or Microsoft camps, who competed with each other for the so-called Teraflop. Both guys shouted how the PS4 has 1.8 TFLOPS, and the old Xbox One only has 1.3, Which, of course, suggested many other possibilities on the part of Sony equipment. When, after 7 years, we finally had successors on store shelves, the debate erupted again, this time with double force.
After all, the XSX offers up to 12 TFLOPS and the PS5 “only” 10.28, which will greatly affect the graphics of published games and will clearly prove that the Xbox Series X is more powerful, right? Well…not necessarily. Yes, more power means more possibilities, but the time is near when we’re not arguing about the original solution, because in its modern form, it’s likely to be gone forever. It will be replaced by NVIDIA DLSS or AMD FSR and none of us will be able to notice it. Personally, I’m very happy with this turn of events, but to know what’s going on, let’s move on to discussing the above image rendering techniques.
The two technologies are basically the same. The game image is displayed at a lower resolution by default and then resized to the target value. This means that when a 4K production is launched, it actually works at 1440p or even lower, depending on the selected profile. Today DLSS (or Deep Learning Super Sampling) has been improved so that we can select a preset that works for us, and choose between quality, performance, and an optimal balance. It is a very useful solution, adapted to the power of our computers. What’s especially important is that scaling is virtually non-invasive, virtually invisible to the player, and impacts often knock them out of the insoles.
NVIDIA prides itself on the fact that, thanks to the use of deep learning mechanisms, image quality and sharpness sometimes exceed what we get at the original resolution. All technologies are based on artificial intelligence. It learns the appearance and position of pixels in a particular production, trying to understand which elements should fill in the “empty” spaces when resolution drops. Remember that when displaying a 1080p image, the graphics card should display 2.07 million pixels, but after switching to 4K, those pixels are close to 8.2 million. Simply put, DLSS or AMD FSR must know how to display the 6 million missing points. On the web you will find many different examples of exactly how they work and how amazing results can be achieved.
In the case of NVIDIA, there are currently over 60 titles that support DLSS, and there are more on the horizon. One of the latest products to have the support of this revolutionary technology is the award-winning Doom Eternal. While it’s not new for PC gamers, since owners of RTX series graphics systems can use DLSS for months, AMD FSR will debut on consoles at any moment and could absolutely revolutionize in terms of saving computing power. Both image rendering technologies offer, with absolutely no or minimal quality loss, a performance increase of up to tens of percent. Often on a PC, we can see that by turning on ray tracing and DLSS, we get 40 or even 50 frames per second. The image quality is literally the same and it is impossible to notice any differences.
Will this completely change the concept of native resolution on consoles?
Today there are still fans who are particularly interested in him. Does the production work in native 4K resolution? And if not, is 1440p really downgraded to 4K on a big screen TV? Personally, I don’t notice any greater blur when focusing on pleasure, but at 1080p I can already notice that the sharpness of the image is very weak. We get used to the good stuff quickly and easily. This is why I appreciate the tremendous effort that went into developing AMD FSR technology. The Reds wanted their answer to NVIDIA DLSS and it looks like they’ve finally brought a technology to market that has the opportunity to dramatically change the perception of displaying images in native resolution.
The results presented by the company look incredibly attractive, but the tests clearly show that AMD still needs to work on its technology a little more, especially in the context of image clarity in motion or when viewing dynamic scenes. This doesn’t change the fact that in the Godfall example on a PC, at 4K resolution with ray tracing enabled, the PC with the RX 6800 XT graphics card produces an average of about 50 frames per second. Already when using the more stressful FSR mode, we get about 30 frames with the sharpness of the image reduced almost to zero. And if we use the performance mode, which significantly blurs the image (makes it at a very low resolution), then instead of 50 frames we get closer to 150. The jump is amazing.
Now imagine how it could work on consoles – there we are practically limited to two bands – 30 or 60 frames per second. Some developers have ambitions to deliver 120 frames per second at the expense of resolution and number of details. The FSR should significantly expand the range of options available. Suppose, purely in theory, the XSX Perfect Dark developers aim for 60fps, but now have to compromise between resolution and detail. The game, in pure theory, runs at 40 frames per second. Thanks to the skillful use of FSR, you no longer have to combine low detail. You can get more, because suddenly, even in this first mode, we gain about 25-30 frames per second, which allows you to achieve 60 frames per second without any problems, transferring the excess power to other elements.
I am very excited about the possibilities that the above technologies bring with them. When using resource-intensive ray tracing, DLSS or AMD FSR are literally useful. Other than that, we would have waited a long time for efficient enough graphics cards, and it would have been difficult for us to see more exciting effects on consoles. FSR may affect the potential premiere of PRO models, as the technology will be available on PS5 and Xbox Series X/S. And if in a year or two these consoles will continually get better products and give advice, will Sony or Microsoft decide to offer stronger variants? Feel free to discuss in the comments!