Support NeoGAF

[Karim]

I dunno g, if only there was some kind of watermark on the picture with the source.



Because it uses an AI heuristic to accumulate the samples rather than a hardcoded algorithm like FSR2 etc? Just like any other AI temporal upscaler like DLSS/FSR4/XeSS.

How upscaling the IQ with the previous frame can be considered the same thing of on AI neural network reconstruction. It's like to say screen space reflections is raytracing reflection. Seems an oversemplification. Though I have to admit I never thought such upscalers were considered simple TAAU, my fault.