The market for graphics cards has undergone significant development in recent years, and today it is clear that NVIDIA has stopped limiting itself to focusing its efforts on brute force, betting generally on more intensive use of AI (artificial intelligence), for implementation new technologies and also for improving efficiency.

We may like it more or less, but we are at a point where raw power under rasterization stopped being relevant a few years ago. No, I don’t mean that it doesn’t matter anymore, nothing is further from reality, I’m just saying that We can no longer judge a graphics card by raw performance alonenor can we put in the same bag two new technologies that are not at the same level, even if they are similar.

NVIDIA has been a big pioneer in this regard and it shows AMD and Intel followed in their footsteps it is clear that those in green succeeded in the end. There was a lot of controversy when the first generation of DLSS came out, and when DLSS 2 came out, which was one of those really cool new technologies, there was also criticism and rejection. Finally AMD released FSR and later FSR2 and we know sooner or later they will release FSR3 with frame interpolation.

The new technologies were very positive for the world of computer games, as they made it possible to play with less powerful graphics cards and with lower consumption in resolutions and graphic qualities that would otherwise not be possible. It also opened the door to ray tracing, as it made it possible to compensate for the performance impact that such technology has.

NVIDIA says so without artificial intelligence, ray tracing, and more specifically path tracing, would indeed be impossible in the world of video games, and the truth is, he’s right. Intelligent scaling and image reconstruction along with AI frame generation will take a game like Cyberpunk 2077 from work to in-between 16 and 34 frames per second on a GeForce RTX 4090, set to 4K, ultra quality and path tracking to move at speeds between 115 and 138 frames per second.

Achieving the raw power necessary to achieve this level of performance without AI would be impossible today and present complications. in terms of consumption, production costs and space at the silicon level they are so large that it would remain impossible for years. Despite all this, it is very clear to me that we need to give AI and new technologies the importance they deserve and stop limiting ourselves to the tunnel vision that some force around brute force. Now it’s your turn read in the comments.