NVIDIA’s graphics drivers for Tensor Core ‘RTX’ GPUs will include AI optimizations.
Artificial intelligence, or AI, has become a popular buzzword for technological advances, particularly in the discovery of new scientific discoveries and medicines, as well as in software and computer components. It’s not surprising that PC components, specifically NVIDIA’s Game Ready Drivers, could pave the way for greater involvement in graphics. Please keep in mind that this is a rumour and that no claims have been confirmed by NVIDIA.
The NVIDIA Game Ready Drivers library has helped improve performance and quality in hundreds of next-gen games as well as some older titles (Quake RTX and Portal RTX), demonstrating that NVIDIA is standing by its claims made last year about their artificial intelligence strategy. In this regard, AI could aid in performance and other related instructions and tasks, allowing games to run more smoothly with less graphical interference, removing some, if not all, of the load from the CPU, and processing that information better or more efficiently.
Make it better by working harder. Doing it faster strengthens us. Work is never over, more than ever, hour after hour.
— Daft Punk’s Harder, Better, Faster, Stronger
CapFrameX tweeted this morning about upcoming drivers that would optimise the NVIDIA Game Ready Drivers with AI enhancements and would be released this year, possibly sooner than expected. It should be noted that the Twitter leaker posted this as a rumour and that it should be treated as a future possibility. It remains to be seen whether these improvements will be delivered via a separate suite or integrated into existing RTX technologies such as DLSS, which rely heavily on AI Tensor cores. NVIDIA did promise to add a variety of improvements and deliver a major update to DLSS by the end of this quarter.
AMD is also working to incorporate AI into its upcoming FSR 3 technology, and their recent Ryzen 7040 ‘Phoenix’ APUs use Xilinx-based AI technology for faster ML and DNN capabilities. This new driver advancement would make use of AMD Radeon Super Resolution, which improves specific latency enhancements and upscaling across the entire system rather than just specifying the overhead to one section of the computer.
With the introduction of its tensor core design, NVIDIA has been at the forefront of AI innovation, and these cores are likely to play a critical role if this news is true. Again, because this is highly speculative, it is unclear how any of this would work or what hardware would be required.