Although RT Cores could sit on the sidelines for a majority of the games (at least for the time being), Tensor Cores play a much larger factor in the performance of Nvidia's new Turing graphics cards.
In fact, they seem to do as many calculations as traditional floating comma and integer pipelines traditionally found in GPU architectures.
If there is something that AI is good at, it is profound learning, which in turn is spectacular in working with images and especially anti-aliasing, which is anything but essential in the world of high-resolution gaming. Anti-aliasing is essential because it removes the jagged edges and the kick effect caused by square pixels.
Technically speaking, anti-aliasing involves identifying two comparable pixels and determining whether they should be combined for a more uniform image. Typically, GPUs so far would handle the process by copying the same values in order, but Tensor kernels can assume a full array of values and compute them all at once.
Nvidia claims that with this method Turing is eight times faster in processing anti-aliasing than Pascal.