NVIDIA DLSS 3 and GeForce RTX 4090: A technical look at NVIDIA’s new technology
September 29, 2022
0
Digital Foundry has published an interesting and complete video review where they test NVIDIA DLSS 3, a major update to the famous image resizing and reconstruction technology that
Digital Foundry has published an interesting and complete video review where they test NVIDIA DLSS 3, a major update to the famous image resizing and reconstruction technology that is now able to generate images using artificial intelligencesomething that, as we told you back then, helps improve performance even in situations where the CPU acts as a bottleneck.
This is one of the most important keys to NVIDIA DLSS 3, because when we activate this technology, the base resolution from which the game is rendered is significantly reduced. This improves performance by reducing the number of pixels in the scenebut at the same time increases CPU impact and may end up creating a bottleneck.
It’s not hard to understand when we activate the technology NVIDIA DLSS 2 or DLSS 3 in performance mode only half of the pixels are rendered horizontally and vertically at the target resolution, meaning that If we want to play in 4K, the basic resolution will be 1920 x 1080 pixels.. Starting at this resolution, the CPU has a lot of influence because the GPU doesn’t really work with more than 8 million pixels, but only moves a little more than 2 million pixels.
To improve performance it was not possible to lower the resolution further plastering of the base and improvement of the rescaling and reconstruction process, from that the bottleneck at the CPU level would become more and more serious. That’s why, with the arrival of NVIDIA DLSS 3, the green giant introduced artificial intelligence-based frame generation, a technology that uses two sequential frames to create an intermediate frame without having to resort to the CPU and without starting from a certain number of frames. pixels.
the whole frame it is created on the GPU without a prior pixel baseand, something made possible by this analysis of two sequential images rendered in the traditional way and combined motion vectors and optical flow field technology. This whole process is accelerated in the Optical Flow Accelerator hardware present in the GeForce RTX 40, and it is an exclusive technology of this generation, because according to NVIDIA, the said hardware is twice as fast and much more accurate than the version present in the GeForce. RTX 30.
NVIDIA DLSS 3 can boost gaming performance by 5x
With image generation present in NV IDIA DLSS 3, the green giant managed to overcome one of the most important limitations which occurred when downscaling the base resolution and rebuilding to a higher one. I’m talking about the bottleneck at the CPU level that I explained to you earlier, and that explains why NVIDIA hit the mark with the introduction of frame generation through artificial intelligence. In the video, Digital Foundry hints at what NVIDIA DLSS 3 technology has to offer up to 500% improvement versus native resolution.
When generating an intermediate frame noticeably improves the flow of the game and this is clearly visible, you only have to see the beginning part, in which Digital Foundry shows us animations of Spider-Man with and without frame generation activated. The difference is substantial and in terms of performance the jump is huge, so much so that you can easily double the performance in this title.
For example, in Cyberpunk 2077 with maximum quality and ray tracing set to “psycho”, we see that GeForce RTX 4090 outperforms GeForce RTX 3090 Ti by 248%. The former used NVIDIA DLSS 3 and frame generation, while the latter used NVIDIA DLSS 2. If we just compare the performance difference between native resolution and DLSS 3, we see that improves performance by up to 399% also in Cyberpunk 2077 with the same setup, while the improvement that DLSS 2 gets is 261%.
As we said, NVIDIA DLSS 3 helps improve performance when a CPU-constrained situation occurs, but if we use a low-powered CPU we will have stuttering problemsi.e. desynchronization when generating frames, and this can completely spoil the gaming experience for us, so we have to take this into account. It won’t help us to mount a GeForce RTX 4090 next to a processor that can’t properly feed it.
in RTX portal is where we see the most noticeable increase in performance, and that is that with NVIDIA DLSS 3 and frame generation we can achieve up to 560% improvement vs native 4K resolution using GeForce RTX 4090 in both cases. Compared to DLSS 2, we see that DLSS 3 triples the performance in said game.
Image generation: what value it provides, what effects it creates and what image quality it offers
Digital Foundry also explains why we need frame generation. The answer seems simple a priori because it allows us to improve performance and because deep down it is an almost natural evolutionary step from the foundation laid by DLSS 2, but there are other things to consider. One of them is the improvement in fluidity of animations, and we have another booming high refresh rate screenswhich require very high frames per second to develop their full potential.
A rasterized game with a “traditional” technical base is currently not a challenge even in 4K thanks to DLSS 2, but if we introduce ray tracing into the equation, things change. That’s where DLSS 3 comes in, and that’s where it makes a huge difference. On the other hand, it also allows us to improve performance even in scenarios where there is a bottleneck at the CPU level, a problem that is very common in games today.
NVIDIA DLSS 3 and frame generation represent uA way to optimize, go beyond brute force and overcome walls that it will squeeze out the GPU footprint, and from what I’ve seen so far, it’s pretty clear that NVIDIA hit the jackpot with this approach. In the final part of the video we can see the analysis of the latency problem, an important problem that because this frame generation process increases the latency.
We have the technology to solve this. NVIDIA Reflex, which is automatically activated when NVIDIA DLSS 3 is running. Thanks to this technology, the latency in Portal RTX with DLSS 3 activated is only 56 ms, which means that it is practically on par with the latency recorded by the title with DLSS 2 and clearly exceeds 95 ms that we have in native resolution. In Cyberpunk 2077, the latency with DLSS 3 activated is 54ms, a figure that exceeds the 62ms of the native mode, although it is below the 31ms of DLSS 2.
In the final part of the video we have a very important problem, the quality of these generated images. According to Digital Foundry, NVIDIA DLSS 3 it does a good job and it is confirmed that it is not a simple frame interpolation technology, as it offers superior image quality and results compared to this type of solution. In the video, we can see the problems that Adobe Pixel Motion gives, for example. Does this mean it is perfect technology? No, it also has bugs and artifacts, but the end result is overall very good and it’s clear that NVIDIA will improve DLSS 3 with future updates.
We can confirm that too consistency in image rendering when NVIDIA DLSS 3 is enabled is goodso we won’t have anything to worry about in this regard if we accompany the GeForce RTX 4090 with a powerful enough processor to “feed” it.
Alice Smith is a seasoned journalist and writer for Div Bracket. She has a keen sense of what’s important and is always on top of the latest trends. Alice provides in-depth coverage of the most talked-about news stories, delivering insightful and thought-provoking articles that keep her readers informed and engaged.