Render Scale & DLSS Resolution Calculator
Calculate internal render resolutions, shading savings, and GPU multipliers for NVIDIA DLSS, AMD FSR, Intel XeSS, and DLDSR supersampling.
DLSS, FSR & XeSS Scaling Reference Matrix
Industry Standards| Scaling Mode | Scale Factor | Render Ratio | Target (4K Output) | Pixel Savings | Action |
|---|---|---|---|---|---|
| DLSS Quality (66.7%) | 1.50� | 66.7% | 2560 � 1440 (1440p) | -55.6% Pixels | |
| DLSS Balanced (58.0%) | 1.72� | 58.0% | 2227 � 1253 | -66.4% Pixels | |
| DLSS Performance (50.0%) | 2.00� | 50.0% | 1920 � 1080 (1080p) | -75.0% Pixels | |
| DLSS Ultra Performance (33.3%) | 3.00� | 33.3% | 1280 � 720 (720p) | -88.9% Pixels | |
| DLDSR 1.78x Supersampling | 1.33� | 133.4% | 5120 � 2880 (5K) | +78.0% Workload | |
| DLDSR 2.25x Supersampling | 1.50� | 150.0% | 5760 � 3240 (6K) | +125.0% Workload |
The Engineering Guide to DLSS, FSR, and Temporal Super-Resolution
Technical analysis of temporal reconstruction math, internal render resolution scaling, shader savings, and AI-driven anti-aliasing.
How Temporal Super Resolution Works (DLSS vs. FSR vs. XeSS)
Modern real-time rendering pipelines (ray tracing, path tracing, volumetric lighting) place massive demands on graphics processing units (GPUs). Real-time reconstruction technologies decouple internal game engine rendering resolution from the target display presentation resolution.
Rather than shading all pixels natively (such as 8.29 million pixels for 4K UHD), the game engine renders geometry, textures, and lighting passes at a lower internal resolution. An upscaling pipeline then reconstructs high-frequency edge detail and sub-pixel stability using temporal motion vectors, jittered depth buffers, and neural networks:
- NVIDIA DLSS (Deep Learning Super Sampling): Uses hardware Tensor cores and temporal autoencoders trained on ground-truth 16K offline renders to predict and reconstruct missing details.
- AMD FSR (FidelityFX Super Resolution): A non-AI spatial and temporal pipeline utilizing Lanczos sharpening and history-clamping algorithms, running across open GPU architectures.
- Intel XeSS (Xe Super Sampling): A neural network reconstruction model accelerated by XMX hardware matrix engines on Intel Arc, with DP4a fallback instructions for competing GPUs.
Internal Render Resolution Math & Pixel Savings Formula
The internal render resolution is calculated by multiplying the target display width and height by the linear scale factor:
Render Height = Round(Target Height � (Scale Percentage / 100))
Total Render Pixels = Render Width � Render Height
Pixel Shading Workload = (Total Render Pixels / Total Target Pixels) � 100
Shading Savings (%) = 100% - Pixel Shading Workload (%)
Because pixel counts scale quadratically along both axes (W � H), a linear reduction to 66.7% (Quality mode) reduces the actual GPU pixel shader payload by 55.6%:
4K DLSS Quality (2560�1440) = 3,686,400 pixels (44.4% load → 55.6% savings)
4K DLSS Performance (1920�1080) = 2,073,600 pixels (25.0% load → 75.0% savings)
DLDSR: Deep Learning Dynamic Super Resolution
While DLSS downscales internal rendering to boost performance on high-resolution displays, DLDSR (Deep Learning Dynamic Super Resolution) performs the inverse. It renders the game engine at a higher virtual canvas (such as 1.78x or 2.25x native pixel density) and downsamples the supersampled frame back to native monitor pixels. This resolves sub-pixel geometry aliasing, eliminates shimmer on fine alpha textures (such as vegetation and fences), and delivers cleaner anti-aliasing without the performance penalty of traditional brute-force SSAA.
Internal Render Scaling Calculations
Formula: Render Width = Target Width � 0.6667
- Internal Res: 3840 � 0.6667 = 2560 px, 2160 � 0.6667 = 1440 px
- Render Resolution: 2560 � 1440 (1440p)
- Target vs Render: 8.29 MP native → 3.69 MP rendered
- Pixel Savings: 55.6% fewer pixels shaded
Formula: Render Width = Target Width � 0.5000
- Internal Res: 2560 � 0.50 = 1280 px, 1440 � 0.50 = 720 px
- Render Resolution: 1280 � 720 (720p HD)
- Target vs Render: 3.69 MP native → 0.92 MP rendered
- Pixel Savings: 75.0% fewer pixels shaded
Frequently Asked Questions
What is DLSS render resolution?
DLSS renders a video game at a lower internal resolution (e.g., 1080p) to boost framerates, then uses AI Tensor cores to reconstruct the image to match your native display resolution (e.g., 4K).
What are the standard DLSS scale factors?
DLSS Quality mode renders at 66.7% per axis (1.5x scale), Balanced renders at 58.0% (1.72x scale), Performance renders at 50.0% (2.0x scale), and Ultra Performance renders at 33.3% (3.0x scale).
How much GPU shading load does DLSS save?
DLSS Quality reduces total pixels shaded by 55.6%, Balanced reduces pixel load by 66.4%, Performance cuts load by 75.0%, and Ultra Performance reduces shaded pixels by 88.9%.
What resolution does 4K DLSS Quality render at?
At 4K UHD (3840�2160), DLSS Quality renders at 2560�1440 (1440p QHD), dropping per-frame rasterization from 8.29 megapixels down to 3.69 megapixels.
What resolution does 1440p DLSS Quality render at?
At 1440p (2560�1440), DLSS Quality renders at 1707�960 (960p), cutting pixel shading workload by 55.6% from 3.69 megapixels to 1.64 megapixels.
What is the difference between DLSS, FSR, and XeSS?
DLSS uses proprietary NVIDIA Tensor cores with neural temporal autoencoders. AMD FSR uses spatial/temporal Lanczos filtering executable on any GPU architecture. Intel XeSS uses XMX AI engines on Intel Arc GPUs and fallback DP4a vector instructions on AMD/NVIDIA cards.
What is DLDSR supersampling?
Deep Learning Dynamic Super Resolution (DLDSR) renders games at a higher internal resolution (e.g., 1.78x or 2.25x native pixels) and downsamples back to your monitor's native panel, producing superior geometric anti-aliasing.
Does DLSS 3 Frame Generation change the internal render resolution?
No. Frame Generation (DLSS 3) operates as a post-render interpolation pass using Optical Flow Accelerators. Internal render resolution is determined strictly by the DLSS Super Resolution mode selected.
Explore Display Engineering Tools
Explore related calculators across the display technology suite.
Calculate diagonal physical dimensions, aspect ratios, total megapixels, and PPI sharpness.
Analyze pixel pitch (mm), visual acuity limits, and Retina distance thresholds.
Calculate GCD aspect ratios, decimal multipliers, and solve missing dimensions.
Check uncompressed data rates and DSC requirements for HDMI 2.1, DisplayPort 2.1, and USB4.
Convert Hz to frame render budgets (ms) and calculate MPRT backlight strobe persistence.
Analyze 1000R, 1500R, and 1800R chord widths, sagitta depth, and horizontal wrap angles.
Simulate retro console scaling, pillarbox/letterbox areas, and display utilization.
Calculate composite canvas resolution, desk width clearance, and sim-rig FOV angles.
Find optimal sitting distances according to THX 36�, SMPTE 30�, and 20/20 acuity limits.