Finally, we can now once again interactively compare the three image acceleration methods. You can use the mouse wheel to zoom in and out, hold down the left mouse button to move the image position, or make settings in the menus. The help function can be found behind the question mark in the upper right corner. If all this is still too small for you despite the zoom, you can open the whole thing as a separate full screen in an extra page by clicking on the link below the comparison. There, too, the operation is identical.
Performance Setting: DLSS 3.0 vs. FSR 2.1 vs. XeSS
Quality Setting: DLSS 3.0 vs. FSR 2.1 vs. XeSS
Summary and conclusion
NVIDIA’s implementation of Frame Generation works with the in-house DLSS, but also with FSR and XeSS from the two competitors. There are hardly any differences between DLSS and FSR in terms of performance, only XeSS lags behind a bit. However, this is also due to the fact that the process in this game does not offer such fast frame times. But it still works. The problem with the latencies, if it is one for the respective user, will be discussed in more detail in the additional article about the launch of the GeForce RTX 4080. That Frame Generation works separately from the individual image acceleration processes was in the focus today. Besides, redundancies are rather boring.
If AMD manages to get a similarly good implementation with the additionally generated frames and also offers this to older or even third-party cards, the chances for a real competition are not bad. But FSR 3.0 has to be released in the first place and it also remains to be seen which way AMD will really go in the end. Do you also push a calculated frame in between or do you perhaps find an alternative approach that doesn’t have to be worse? Let’s be surprised, because today’s article is just a kind of appetizer.