NVIDIA's RTX NTC (Neural Texture Compression) has seen massive improvements with Microsoft's DirectX Cooperative Vector, and the latest tests show a gigantic drop in VRAM consumption.
Demo Rendering Tests Show That NVIDIA's NTC Could Bring In Massive Performance Improvements With DirectX Raytracing 1.2
Graphical advancements aren't just limited to GPU hardware; this has been proven in the past with upscaling technologies like NVIDIA's DLSS. Now, based on a post by @opinali, who has managed to test out NVIDIA's Neural Texture Compression, combined with Microsoft's latest DirectX Raytracing 1.2, performance levels are pretty impressive. Based on testing done by the user, it is revealed that under rendering scenarios, notably the iconic shoe-render, the user noted a massive drop in VRAM usage with NTC and DirectX Raytracing (DXR) 1.2 combined.
Based on what we know, NVIDIA's Neural Text Compression uses neural networks to compress and decompress game textures, allowing a reduction in size with minimal impact to quality. Apart from this, one of the key qualities of the Microsoft DXR 1.2 update is the inclusion of Cooperative Vectors, which is a method to allow GPU shaders to work together on small matrix or vector operations. Now, when you combine NTC and Cooperative Vectors, you get a compression/decompression mechanism that runs efficiently within standard game shaders via DX12, which has reduced VRAM usage.
The user's testing reveals that by enabling Cooperative Vectors and NTC together under "Default" mode, the texture renders at 2,350 FPS; however, when you disable it completely (DP4A), the rendering performance drops to 1,030 FPS, marking almost an 80% difference. The VRAM capacities have also seen a massive difference, and interestingly, NTC is limited to NVIDIA's GPUs right now, since Intel/AMD alternatives don't have their neural rendering kits refined up till now.
It's important to note that NTC is currently accessible with the newest 590.26 preview drivers, which also bring in NVIDIA Smooth Motion technology, but experimenting with such drivers would cause performance disruptions. When you combine these refinements, the next-generation GPUs will certainly have a much more profound impact on performance with software capabilities.
About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.
Follow Wccftech on Google to get more of our news coverage in your feeds.
