Nvidia's DLSS, or Deep Learning Super Sampling, is a revolutionary feature that has significantly transformed PC gaming since its introduction in 2019. This technology not only enhances performance but also extends the life and value of Nvidia's RTX graphics cards, particularly for games that support it. As the system has evolved through various updates, it has become more effective and has introduced differentiating features across Nvidia's RTX generations. This guide will delve into what DLSS is, how it functions, its variations, and why it's relevant to gamers, even if you're not currently using an Nvidia GPU.
*Additional contributions by Matthew S. Smith.*
What Is DLSS?
Nvidia DLSS, or Deep Learning Super Sampling, is a proprietary technology designed to boost both performance and image quality in video games. The term "Super Sampling" refers to its capability to upscale games to higher resolutions using a neural network trained on extensive gameplay data. This allows for a performance boost without the usual impact of manually setting a higher resolution in-game.
Originally focused on upscaling, DLSS has expanded to include several other systems that enhance image quality. These include DLSS Ray Reconstruction, which uses AI to improve lighting and shadow quality; DLSS Frame Generation and Multi Frame Generation, which use AI to insert additional frames to boost FPS; and DLAA (Deep Learning Anti-Aliasing), which applies AI-enhanced anti-aliasing for superior graphics compared to native resolution.
The most recognized feature of DLSS is Super Resolution, particularly beneficial when combined with ray tracing. In games that support DLSS, you can find settings like Ultra Performance, Performance, Balanced, and Quality within the graphics menu. For instance, in Cyberpunk 2077, selecting 4K resolution with DLSS Quality mode means the game renders at 1440p, which is easier to run, and DLSS upscales it to 4K, resulting in a higher frame rate than achievable at native 4K.
DLSS's neural rendering differs significantly from older techniques like checkerboard rendering. It can add details not visible at native resolution and preserve details lost in other upscaling methods. However, it may introduce artifacts like "bubbling" shadows or flickering lines, though these have been greatly reduced with DLSS 4.
The Generational Leap: DLSS 3 to DLSS 4
With the RTX 50-series, Nvidia introduced DLSS 4, which significantly enhances the AI model used in the system. DLSS 3 and its 3.5 update, which introduced frame generation, utilized a Convolutional Neural Network (CNN) trained on vast video game datasets. This model analyzed scenes, spatial relationships, and key objects.
DLSS 4, however, shifts to a Transformer model, or TNN, which can analyze twice as many parameters, offering a deeper understanding of scenes. This model interprets inputs more sophisticatedly, including long-range patterns, leading to better results across all DLSS features. DLSS Super Sampling and DLSS Ray Reconstruction in DLSS 4 retain more fine details, resulting in sharper gameplay and fewer artifacts like bubbling shadows and flickering lines.
The TNN model also enhances frame generation. While DLSS 3.5 inserted one frame between two natively rendered frames, DLSS 4 can generate four artificial frames per rendered frame, known as DLSS Multi Frame Generation. This can significantly increase frame rates. To address concerns about input lag, Nvidia integrates Nvidia Reflex 2.0, which reduces latency to maintain responsiveness.
Despite its advancements, DLSS Frame Generation can still produce minor ghosting behind moving objects, especially at higher settings. Nvidia allows users to adjust frame generation levels and recommends matching settings to your monitor's refresh rate to avoid issues like screen tearing.
Even without an RTX 50-series, you can benefit from the new Transformer model for DLSS Super Resolution and DLSS Ray Reconstruction using the Nvidia App. The app also enables DLSS Ultra Performance mode and DLAA for games that don't natively support these options.
Why Does DLSS Matter for Gaming?
DLSS is a game-changer for PC gaming, particularly for those with mid-range or lower-performance Nvidia GPUs. It allows access to higher graphics settings and resolutions that would otherwise be unattainable. Additionally, DLSS extends the lifespan of your GPU by maintaining playable frame rates through adjusted settings or performance modes, making it a cost-effective solution for gamers on a budget.
DLSS has also spurred competition, with AMD and Intel introducing their own upscaling technologies, AMD FidelityFX Super Resolution (FSR) and Intel Xe Super Sampling (XeSS). While Nvidia's DLSS has set a high standard with its image quality and frame generation capabilities, the competition has driven innovation across the industry, ultimately benefiting gamers.
Nvidia DLSS vs. AMD FSR vs. Intel XeSS
Nvidia's DLSS faces competition from AMD's FidelityFX Super Resolution (FSR) and Intel's Xe Super Sampling (XeSS). DLSS 4 offers superior image quality and multi-frame generation with low input latency, giving it an edge over its rivals. While AMD and Intel's technologies also provide upscaling and frame generation, Nvidia's machine learning capabilities lead to crisper, more consistent images with fewer artifacts.
It's worth noting that DLSS is exclusive to Nvidia GPUs and requires game developer implementation, unlike AMD FSR. Although the number of supported games has grown significantly, not all games support DLSS, FSR, or XeSS by default.
Conclusion
Nvidia DLSS has revolutionized the gaming industry and continues to improve. It's a testament to Nvidia's commitment to enhancing gaming experiences and extending GPU longevity. While not perfect, DLSS can significantly impact your gaming experience at its best.
However, DLSS is no longer the only player in the field, with AMD and Intel offering competitive technologies. As with any PC gaming hardware, it's crucial to weigh the cost and features of your GPU against the games you play to find the best value for your needs.
