What Is DLSS, FSR, and XeSS? Upscaling Explained
AI upscaling lets you get higher frame rates without sacrificing image quality. Here is how DLSS, FSR, and XeSS work and which one your GPU supports.
BestElectronicsReviewed.com is a participant in the Amazon Services LLC Associates Program. We may earn a commission from qualifying purchases made through links on this page, at no extra cost to you.
Upscaling technology has become one of the most important innovations in gaming. Instead of rendering a game at your display's native resolution — which demands enormous GPU power — upscaling renders at a lower resolution and uses algorithms or AI to reconstruct a sharp, detailed image. The result is dramatically higher frame rates with minimal visual compromise. Three technologies dominate this space: NVIDIA's DLSS, AMD's FSR, and Intel's XeSS.
DLSS: The AI Pioneer
Deep Learning Super Sampling, or DLSS, is NVIDIA's upscaling technology available exclusively on RTX GPUs. It uses dedicated AI accelerator hardware (Tensor Cores) to upscale images, and it has been trained on thousands of game frames to produce results that often look as good as — or sometimes better than — native resolution rendering.
DLSS works by rendering the game at a lower internal resolution (for example, 1080p) and then using an AI model to reconstruct a 4K image. The AI model analyzes motion vectors and temporal data from previous frames to fill in detail that the lower render resolution missed. The result is an image that looks remarkably close to native 4K while running at frame rates you would expect from 1080p rendering.
DLSS 3 added Frame Generation, which creates entirely new intermediate frames between real rendered frames. This effectively doubles perceived frame rates. A game rendering at 60 real fps can display at 120 fps with generated frames filling the gaps. Frame Generation does add a small amount of input latency (about half a frame), but for single-player games, the visual smoothness improvement is worth the trade-off.
DLSS 3.5 introduced Ray Reconstruction, which uses AI to denoise ray-traced effects more efficiently than traditional methods. This means ray tracing runs faster and looks better simultaneously — a genuinely impressive technical achievement.
An RTX 4070 with DLSS 3 can effectively deliver 4K performance from a mid-range card, making it one of the best value propositions in PC gaming.
FSR: AMD's Open Alternative
FidelityFX Super Resolution, or FSR, is AMD's answer to DLSS. Unlike DLSS, FSR is open source and works on virtually any GPU — AMD, NVIDIA, and Intel. This broad compatibility is its biggest advantage.
FSR has gone through significant evolution. FSR 1.0 was a simple spatial upscaler that sharpened a lower-resolution image. It was noticeably inferior to DLSS. FSR 2.0 and beyond added temporal upscaling — using data from previous frames to reconstruct detail, similar in concept to DLSS but without dedicated AI hardware.
FSR 3.0 added Fluid Motion Frames, AMD's version of frame generation. Like DLSS 3, it inserts AI-generated frames between real frames to boost perceived smoothness. FSR 3's frame generation is available on both AMD and NVIDIA GPUs, and early implementations show competitive results with DLSS 3 in many titles.
The quality gap between FSR and DLSS has narrowed considerably. At "Quality" and "Balanced" presets, FSR 2.1+ produces images that are difficult to distinguish from DLSS in most games. At lower presets (Performance and Ultra Performance), DLSS still holds a visible edge, particularly in preserving fine detail and reducing ghosting artifacts.
XeSS: Intel's Entry
Intel's XeSS (Xe Super Sampling) is the newest contender. Like FSR, it works across GPU vendors, but it provides enhanced quality on Intel Arc GPUs through dedicated XMX AI cores — similar to how DLSS works best on NVIDIA's Tensor Cores.
On Intel Arc GPUs, XeSS uses AI-based upscaling powered by the hardware's matrix acceleration units. On AMD and NVIDIA GPUs, it falls back to a DP4a shader-based path that is less efficient but still functional. Quality on Intel hardware is competitive with DLSS. On non-Intel hardware, quality is roughly comparable to FSR.
XeSS is still building its game library, with fewer supported titles than DLSS or FSR. However, as Intel continues investing in its Arc GPU lineup and developer relations, XeSS support is growing steadily.
Which Should You Use?
The answer depends entirely on your GPU. If you have an NVIDIA RTX GPU, use DLSS whenever available — it provides the best quality and includes frame generation and ray reconstruction features. An RTX 4060 paired with DLSS 3 punches well above its weight class.
If you have an AMD GPU, use FSR. It is available in a rapidly growing number of games and delivers excellent results at Quality and Balanced presets. Frame Generation via FSR 3 is a meaningful bonus.
If a game supports both DLSS and FSR, try both and see which looks better to your eyes. In some titles, DLSS has a clear edge. In others, they are indistinguishable.
The Bigger Picture
Upscaling has fundamentally changed the economics of PC gaming. A mid-range GPU from 2026 with DLSS or FSR delivers visual quality that would have required a flagship card three years ago. The technology is only getting better — future versions promise even higher quality reconstruction and smarter frame generation. When evaluating a GPU purchase, upscaling support should be a primary consideration alongside raw rasterization performance.
As an Amazon Associate, BestElectronicsReviewed earns from qualifying purchases.
Recommended Products
Top picks from our buying guides
Related Articles
How Zigbee, Z-Wave, and Thread Actually Differ
Smart home protocols are confusing. Here's a plain-English breakdown of the three mesh networking standards and which one matters most in 2026.
ExplainerWhat Is WiFi 7 and Do You Actually Need It? (Today)
What Is WiFi 7 and Do You Actually Need It? (Today) — expert analysis and tested recommendations from BestElectronicsReviewed.
ExplainerHow Robot Vacuum Navigation Works: LiDAR vs Camera vs Gyroscope
The navigation system is the most important feature in a robot vacuum. Here's how LiDAR, camera, and gyroscope systems work, and which cleans best.