Deep Learning Super Sampling or DLSS is a style of video rendering that aims to increase framerates by rendering frames at a lower resolution than what is displayed and then upscales the frames using deep learning, an AI technique, to make them appear as clear as expected at the native resolution. For instance, using DLSS, a game’s frames could be output at 4K resolution after being produced at 1080p resolution to provide sharper image quality than 1080p and enable greater framerates.

Only RTX graphics cards support DLSS, but AMD has unveiled AMD Fidelity FX Super Resolution, its own implementation of the technique (AMD FSR). As long as the required hardware characteristics are there, FSR can run on GPUs from both Nvidia and Intel.

Nvidia claims that DLSS 2.0, the most widely used version, may increase framerates by 200–300%. Fewer games use the original DLSS, although Nvidia claims it can increase framerates “more than 70%”. Even with the greatest graphics cards, DLSS can greatly help when playing games at high resolutions or with ray tracing, both of which can cause framerates to drop noticeably compared to 1080p. It is challenging to distinguish between a game rendered in native 4K and one that was upscaled from 1080p to 4K using DLSS 2.0 (that is, the ‘performance’ mode with 4x upscaling). Although the performance improvements aren’t as significant, it’s nearly hard to notice the difference between DLSS 2.0 in quality mode (i.e., 1440p upscaled to 4K).

Required cards

Although DLSS is fantastic for higher resolutions, there is a catch. DLSS is not accessible for those still running GTX 10 series graphics cards. The RTX 2000 and 3000 series GPUs’ Tensor cores are used by DLSS, making them an absolute must for using the technology. Though the new Tensor cores offer higher per-core performance, DLSS will run better on the 30 series cards, which still outperform the RTX 2000 series. Even though the Steam Hardware Survey indicates that GTX 1060 cards are the most common on the market, it is regrettable that DLSS is not supported on older generation cards. However, this is impossible since 10-series GPUs lack Tensor Cores.

Final Words

The outstanding integration of Nvidia’s DLSS greatly enhances fidelity and performance. Additionally, it gives people who don’t necessarily have the gear for higher settings or better resolutions access to an experience. Additionally, it has been wonderful to see Nvidia integrate the Tensor cores required for DLSS in their most recent low-cost model, the RTX 3050. Giving DLSS support to gamers on a tight budget is a significant advancement and a promising harbinger of things to come.