As early as the launch of the RTX 20 series graphics cards, NVIDIA pioneered the introduction of DLSS deep learning hypersampling, and with the introduction of new generations of graphics cards, the technology is also iteratively optimized. Later, NVIDIA launched DSR dynamic super resolution technology and NIS quality scaling technology, and supported open source, becoming the industry benchmark.
On January 12, Beijing time, NVIDIA officially launched the DLDSR technology, the full English name is Deep Learning Dynamic Super Resolution, which translates to deep learning dynamic super resolution technology.
DLDSR technology is supported by the tensor core of NVIDIA GPU tensor on the hardware, and adds AI capabilities on the basis of DSR, which can bring players a more realistic and smooth picture quality experience. It is expected that both RTX 20 and RTX 30 series graphics cards will support this technology.
DSR technology, DLDSR technology is rendered through high resolution, and then compressed to the screen native resolution output, can greatly enhance the details of the picture, make the edges smoother, less aliasing, but also reduce glare. It is worth noting that DLDSR joins the AI network layer and requires fewer input pixels, so it can output higher frame rates.
In addition, NVIDIA also said that DLDSR technology does not require special optimization of the game, most games support it, and the technology will be directly integrated into the driver.
(Image from the Internet)