You can get more out of graphics with the help of artificial intelligence, the second version already works convincingly – Živě.cz

Deep Learning Super Sampling (DLSS) is Nvidia’s real-time image reconstruction technology. The first version did not dazzle, but the second generation is a big step forward.

Nvidia began working on DLSS because with new technologies such as ray tracing, more realistic physics, augmented and virtual reality, and high-resolution displays, the demands on graphics chip performance increase exponentially. Ray tracing alone is computationally several times more demanding than traditional rendering techniques.

But there was another good reason for the emergence of DLSS. GeForce cards from the RTX family use the same graphics chips as Nvidia computer cards from the professional sphere, which are also used in calculations for artificial intelligence. For this reason, specialized units also received tensor cores for accelerating calculations performed via neural networks for deep learning. And when they’re in the chip, Nvidia wanted to use them on gaming graphics cards as well.