Intel has revealed to the world the new ExtraSS technology to try to undermine Nvidia’s dominance and the premises are interesting.

Technology makes progress that remains hidden until one of two situations occurs: someone behind the scenes manages to get their eyes on some detail and spreads it to the world or the companies directly involved decide that the time has finally come to show to the world what they are working on.

ExtraSS against DLSS, Intel challenges AMD and Nvidia – Sjbeez

Intel’s new technology called ExtraSS was the subject of the second type of announcement. In fact, the company participated in an event held recently in Sydney and the new ExtraSS technology has already been dubbed as revolutionary. This is actually Intel’s version of DLSS 3 and FSR 3 but it doesn’t work exactly the same way. The company immediately admitted that it is still an experimental reality with a series of limitations but the potential is there and just needs to be unleashed.

Because we will soon hear about Intel and ExtraSS

Acronyms like DLSS and FSR may not mean anything to the wider public but for those who work with video games and for the most attentive gamers it is what makes the gaming experience fluid. These are frame generation technologies that allow you to enjoy more beautiful and realistic animations. The technology that unites DLSS and FSE is frame interpolation, i.e. the generation of intermediate frames between those already present. Intel’s idea is to overturn what has been done so far by working not on interpolation but on the extrapolation of frames.

Intel in the Olympus with AMD and Nvidia, ExtraSS arrives – Sjbeez

The differences are in the speed with which the technology is able to produce the new frames. In fact, interpolation needs two frames to calculate the third to be interpolated and to do this it produces a slight lag. The extrapolation imagined by Intel with the ExtraSS system instead works on a single frame, which through motion vectors and spatial data is able to produce the next frame: it then extrapolates the data and uses it to predict the next necessary frame.

A technology which on paper should therefore be much faster than the interpolation given for example by DLSS 3 but, Intel clearly admits, glitches and artefacts can still occur and above all to work on data extrapolation it is necessary that in the frames there is there is a sufficient quantity. On social media where the news of the new Intel technology has spread, a discussion has also reopened that is worth re-proposing every now and then: do we really need realistic games down to the last hair in our nose?

A question for which there are two answers and the factions of the two answers bring contrasting but plausible motivations. Technologies that are able to increase the number of frames and therefore make animations more fluid and realistic and provide images with more details can be useful but, and someone points out, there should be a good number of frames and good quality starting without hoping that Nvidia, AMD or Intel will do some magic.