AMD Radeon RX 7900 XT vs GeForce RTX 4090
VS
AMD Radeon RX 7900 XT
Navi 31
Nvidia GeForce RTX 4090
AD102
We compared two discrete desktop gaming GPUs: the Radeon RX 7900 XT 20 GB with 5376 shaders against the 1 year and 3 months older GeForce RTX 4090 24 GB that utilizes 128 pipelines and 16384 shaders. Here you will find complete details about specs, efficiency, performance tests, and more.
Review
Gaming
Performance in DirectX, OpenCL, and Vulkan games
Workstation
Perf. in 3D modeling, video editing and rendering apps
AI/ML
Capabilities for machine learning and AI-related tasks
Energy Efficiency
Power consumption efficiency in different scenarios
DeviceFacts Final Score
Overall video card score
Value for money (Beta)
Key differences
Reasons to consider the GeForce RTX 4090
- Performs significantly better (up to 71%) in 3DMark Steel Nomad Lite
- 60% higher maximum theoretical performance (82.6 vs 51.5 TFLOPS)
- Shows 25% higher average frame rate in modern games at QHD resolution – 177 vs 142 FPS
- Includes 4 GB more video memory
- Supports Nvidia DLSS 3 technology
- Achieves 55% more points in the GeekBench 6 Compute test (316K vs 203K)
- Has 3x more shading units (16384 vs 5376)
- Has 26% higher memory bandwidth: 1010 vs 800 GB/s
Gaming Performance
Frame rate comparison across popular AAA titles at different resolutionsGames
| Forza Horizon 5 | 235 160 141 107 |
275 199 193 155 |
| The Witcher 3 | 249 225 160 87 |
292 268 191 101 |
| Counter-Strike 2 | 311 222 157 81 |
339 238 188 90 |
| Far Cry 6 | 196 168 151 100 |
203 179 170 104 |
| Hogwarts Legacy | 153 124 98 56 |
200 163 132 74 |
| CoD: Modern Warfare III | 228 214 163 108 |
266 263 219 151 |
| Ghost of Tsushima | 150 126 109 69 |
188 152 137 91 |
| Cyberpunk 2077 | 181 156 108 49 |
226 208 143 66 |
| Shadow of the Tomb Raider | 271 258 189 101 |
302 284 222 118 |
| 1080p High | 219 | 255 |
| 1080p Ultra | 184 | 217 |
| 1440p Ultra | 142 | 177 |
| 4K Ultra | 84 | 106 |
| Margin of Error | Low | Low |
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Time Spy | 27106 | 36299 |
| Solar Bay | 105357 | 187525 |
| Port Royal | 14360 | 26127 |
| Fire Strike | 67227 | 72432 |
| Wild Life Extreme | 48410 | 85218 |
| Night Raid | 176232 | 195813 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 419.8 img/sec | 318.9 img/sec |
| Face Detection | 240.5 img/sec | 222.8 img/sec |
| Horizon Detection | 8.72 Gpixels/sec | 14.2 Gpixels/sec |
| Edge Detection | 12 Gpixels/sec | 21 Gpixels/sec |
| Gaussian Blur | 9.88 Gpixels/sec | 23.9 Gpixels/sec |
| Feature Matching | 1.71 Gpixels/sec | 2.38 Gpixels/sec |
| Stereo Matching | 701.7 Gpixels/sec | 1950 Gpixels/sec |
| Particle Physics | 26631.4 FPS | 48090.4 FPS |
| API | OpenCL | OpenCL |
Cinebench 2024 GPU
Cinebench 2024 GPU
Passmark Graphics
G3D Mark Score
| G2D Mark | 1247 | 1303 |
| DirectX 11 | 341 FPS | 323 FPS |
| DirectX 12 | 112 FPS | 151 FPS |
| GPU Compute | 16878 Ops/s | 25991 Ops/s |
Blender
Blender GPU
Recent User Tests
Radeon RX 7900 XT
| Date | Benchmark | Result |
|---|---|---|
| 📘 2026-01-01 (Mr Scare) | Cinebench 2024 | 11648 |
GeForce RTX 4090
| Date | Benchmark | Result |
|---|---|---|
| 📘 2026-10-06 (Devansh) | Geekbench 6 OpenCL | 394674 |
| 📘 2026-10-06 (Devansh) | Geekbench 7 OpenCL | 338129 |
| 📘 2026-04-03 (afroman420IU) | Cinebench 2024 | 32578 |
Artificial Intelligence Tests
Performance in machine learning and artificial intelligence tasksGeekBench 6 ML
GB6 ML Single Precision
GB6 ML Half Precision
GB6 ML Quantized
| Image Classification (SP) | 16951 | 15600 |
| Image Segmentation (HP) | 23234 | 36657 |
| Image Super Resolution (Q) | 41128 | 48114 |
| Face Detection (HP) | 67615 | 78111 |
| Pose Estimation (Q) | 237026 | 289639 |
| Text Classification (SP) | 4961 | 4074 |
| Machine Translation (HP) | 7043 | 6349 |
| Object Detection (SP) | 18737 | 21186 |
| Depth Estimation (Q) | 67746 | 75599 |
| Style Transfer (SP) | 457919 | 601428 |
| Framework | ONNX | ONNX |
| Backend | DirectML | DirectML |
Specifications
Technical specifications of Radeon RX 7900 XT and GeForce RTX 4090General
| Vendor | Amd | Nvidia |
| Build | Discrete | Discrete |
| Released | December 13, 2023 | September 20, 2022 |
| Launch price (MSRP) | $899 | $1599 |
| Case | Desktop | Desktop |
| Purpose | Gaming | Gaming |
| Segment | High-end | High-end |
| Architecture | RDNA 3.0 | Ada Lovelace |
| GPU Codename | Navi 31 | AD102 |
| Rival Equivalent | - GeForce RTX 4080 | - |
| Successor | - | - GeForce RTX 5090 |
| Recommended CPU | - Intel Core i7 13700K or above | - Intel Core i9 14900K or above |
Desktop GPU rating (11th and 3rd place)
Graphics Processing Unit
| Base Clock | 1387 MHz | 2235 MHz |
| Boost Clock | 2394 MHz | 2520 MHz |
| Shading Units | 5376 | 16384 |
| Texture Mapping Units (TMUs) | 336 | 512 |
| Render Output Units (ROPs) | 192 | 176 |
| Compute Units (Pipelines) | - | 128 |
| Tensor Cores | - | 512 |
| Ray-tracing Cores | 84 | 128 |
| L1 Cache | 256KB per cluster | 128KB per cluster |
| L2 Cache | 6MB shared | 72MB shared |
| Instructions Per Cycle | 4 IPC | 2 IPC |
Raw Performance
| Pixel Fill Rate | 460 GPixel/s | 444 GPixel/s |
| Texture Fill Rate | 804 GTexel/s | 1290 GTexel/s |
FLOPS (FP32)
Differences in GPUs' instructions per clock make TFLOPS values not directly comparable.
Physical
| Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| TGP | 300 W | 450 W |
| Manufacturing | TSMC | TSMC |
| Fabrication Process | 5 nm | 5 nm |
| Die Size | 529 mm² | 609 mm² |
| Transistor Count | 57 billion | 76.3 billion |
| Transistor Density | 107.75 MTr/mm² | 125.29 MTr/mm² |
Memory
| Memory Type | GDDR6 | GDDR6X |
| Memory Size | 20 GB | 24 GB |
| Memory Clock | 2500 MHz | 2625 MHz |
| Effective Memory Speed | 20000 Mbps | 21000 Mbps |
| Bus | 320-bit | 384-bit |
| ECC | No | No |
Memory Bandwidth
API
| DirectX | 12 | 12 |
| Vulkan | 1.3 | 1.3 |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 2.2 | 3.0 |
| CUDA | - | 8.9 |
| Ray Tracing | Yes | Yes |
| DLSS | No | DLSS 3 |
| DisplayPort | 2.1 | 1.4a |
Cast your vote
Total votes: 23