NVIDIA is always on the right track when it comes to improving video games graphically and technologically. The proof is with this novelty.
NVIDIA is seriously working on the subject of artificial intelligence and the American firm is still seeking to develop its tools even further. to revolutionize the graphics of the video game of tomorrow. Some of its new products will be present at the SIGGRAPH 2023 conference to be held from August 6 to 10 in Los Angeles. The aim is to present 20 study points on the subject. And several points are directly or indirectly related to video games.
NVIDIA set to revolutionize video game graphics
Thus one of the research projects of NVIDIA offers a new approach to simulating the hair on a human head. It may seem quite innocuous but on the technical side it is still a perceptible notion in the video game and its representation. Managing more or less 100,000 hairs is a major technological headache for 3D animators. To solve part of the problem, NVIDIA has just found a method that can simulate thousands of hairs in high resolution and in real time. For a brief overview you can watch the video below. You have to admit that it’s stunning and almost disturbing.
Real-time rendering simulates the physics of reflected light through the virtual scene. Recent research from NVIDIA shows how AI models for textures, materials, and volumes can deliver real-time, cinematic-quality photorealistic visuals for video games. A revolution ? Imagine the last Uncharted or The Last of Us with this possibility.
This rather fabulous work is based in particular on the power of graphics chips, another specialty of NVIDIA (that’s good). A good opportunity for the builder to implicitly present the problem and the solution.
AI, a major challenge
In one of the twenty points addressed by the NVIDIA company, there is in particular a texture compression technique which would offer 16 times more texture details without requiring GPU memory. (graph) additional. All thanks to artificial intelligence. To a lesser extent, it is already a bit like the principle of DLSS (deep learning super sampling), which uses AI to generate additional high-quality images. It is easy to imagine the importance of such technology for video games or even animated films. Soon these AI-based technologies will make it possible to offer highly realistic representations of humans who are completely fictitious. Perfect for video game immersion may be a little less for certain professions in the field.
NVIDIA specifies that this research could also find important applications in robotics and for autonomous cars.
