Britannica Sources used in this answer

How does NVIDIA's DLSS improve image quality while lowering rendering costs?

Created from Britannica articles using AI. AI can make mistakes, so verify using Britannica articles.

computer graphics The Utah Teapot, a standard graphics image used to benchmark computer systems.

NVIDIA's Deep Learning Super Sampling (DLSS) improves image quality and performance by utilizing artificial intelligence to reconstruct high-resolution images from lower-resolution inputs.12

Instead of rendering every pixel at a native high resolution—which is computationally expensive—the graphics processing unit renders the scene at a lower resolution. DLSS then employs machine learning and trained neural-network models to infer and fill in visual details. By combining data from motion vectors and previous frames, the technology generates a final output that appears to be at a much higher resolution.13

This process significantly reduces the computational load on the system, allowing for higher frame rates in electronic games without requiring the hardware to perform the full, intensive task of native-resolution rendering.13