About a year ago, NVIDIA officially introduced its proprietary active noise cancellation technology – thanks to the software, the user received a higher quality of audio communication, since unnecessary noises from the outside were “cut off” by machine learning. At the same time, company representatives officially announced that RTX Voice is a unique feature for RTX 20-series GPUs, and to perform sound manipulations, you must have tensor cores, without which it is impossible to improve sound based on machine learning. Today, fortunately, the situation has been corrected – now the software is available for almost all video cards of the company.
At the moment, absolutely anyone can download the RTX Voice application for their GTX graphics card – just have a graphics accelerator that supports NVIDIA driver 410.18 or newer. In fact, all video cards from GeForce GTX 600 up to and including RTX 30 are supported. However, this has already been proven by enthusiasts – without much difficulty, with the help of a simple modification, they were able to run RTX Voice software on fairly old video cards, and even on Windows 7. This experiment was successfully carried out literally a week after the presentation of NVIDIA proprietary technology, when the craftsmen proved that claims about the need for tensor cores are just marketing.
An updated version of the software with support for older video cards is already on the official website.