AMD Radeon RX 9070 XT vs GeForce RTX 5090
VS
AMD Radeon RX 9070 XT
Navi 48
Nvidia GeForce RTX 5090
GB202-300
We compared two discrete desktop gaming GPUs: the Radeon RX 9070 XT 16 GB with 64 pipelines and 4096 shaders against the GeForce RTX 5090 32 GB that utilizes 170 pipelines and 21760 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
NanoReview Final Score
Overall video card score
Value for money (Beta)
Key differences
Reasons to consider the GeForce RTX 5090
- Performs significantly better (up to 2x) in 3DMark Steel Nomad Lite
- 2.2x higher maximum theoretical performance (104.8 vs 48.7 TFLOPS)
- Shows 36% higher average frame rate in modern games at QHD resolution – 207 vs 152 FPS
Gaming Performance
Frame rate comparison across popular AAA titles at different resolutionsGames
| Forza Horizon 5 | 242 169 149 109 |
336 247 233 187 |
| The Witcher 3 | 267 237 167 89 |
364 327 237 123 |
| Counter-Strike 2 | 325 224 163 82 |
363 265 202 99 |
| Far Cry 6 | 198 176 163 104 |
237 202 190 121 |
| Hogwarts Legacy | 162 133 107 58 |
249 199 166 97 |
| CoD: Modern Warfare III | 240 232 174 118 |
334 322 260 184 |
| Ghost of Tsushima | 158 137 116 73 |
221 185 175 113 |
| Cyberpunk 2077 | 190 171 119 53 |
261 233 160 77 |
| Shadow of the Tomb Raider | 280 265 207 107 |
329 302 238 128 |
| 1080p High | 229 | 299 |
| 1080p Ultra | 194 | 254 |
| 1440p Ultra | 152 | 207 |
| 4K Ultra | 88 | 125 |
| Margin of Error | Low | Low |
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Time Spy | 29556 | 46873 |
| Solar Bay | 122220 | 236689 |
| Port Royal | 18363 | 38202 |
| Fire Strike | 66319 | 87569 |
| Wild Life Extreme | 54672 | 108751 |
| Night Raid | 194947 | 206648 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 427.4 img/sec | 333 img/sec |
| Face Detection | 251 img/sec | 298.5 img/sec |
| Horizon Detection | 6.77 Gpixels/sec | 21.7 Gpixels/sec |
| Edge Detection | 9.42 Gpixels/sec | 33.7 Gpixels/sec |
| Gaussian Blur | 9.35 Gpixels/sec | 34.5 Gpixels/sec |
| Feature Matching | 1.42 Gpixels/sec | 2.53 Gpixels/sec |
| Stereo Matching | 651.4 Gpixels/sec | 2850 Gpixels/sec |
| Particle Physics | 23630 FPS | 59146.9 FPS |
| API | OpenCL | OpenCL |
Cinebench 2024 GPU
Cinebench 2024 GPU
Passmark Graphics
G3D Mark Score
| G2D Mark | 1310 | 1410 |
| DirectX 11 | 295 FPS | 340 FPS |
| DirectX 12 | 77 FPS | 179 FPS |
| GPU Compute | 16100 Ops/s | 24383 Ops/s |
Blender
Blender GPU
Recent User Tests
Radeon RX 9070 XT
| Date | Benchmark | Result |
|---|---|---|
| 📘 2026-07-12 (luizmdd) | Cinebench 2024 | 50113 |
| 📘 2026-07-04 -> StevenS | Geekbench 6 OpenCL | 179283 |
| 📘 2026-07-04 -> StevenS | Steel Nomad Light | 27505 |
| 📘 2025-10-15 (adv) | Geekbench 6 OpenCL | 175418 |
GeForce RTX 5090
No benchmark results yet
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) | 18001 | 18948 |
| Image Segmentation (HP) | 25174 | 55109 |
| Image Super Resolution (Q) | 45714 | 58478 |
| Face Detection (HP) | 74776 | 114685 |
| Pose Estimation (Q) | 196137 | 344729 |
| Text Classification (SP) | 5163 | 4464 |
| Machine Translation (HP) | 8677 | 8184 |
| Object Detection (SP) | 19602 | 27802 |
| Depth Estimation (Q) | 74879 | 85572 |
| Style Transfer (SP) | 411302 | 687080 |
| Framework | ONNX | ONNX |
| Backend | DirectML | DirectML |
Specifications
Technical specifications of Radeon RX 9070 XT and GeForce RTX 5090General
| Vendor | Amd | Nvidia |
| Build | Discrete | Discrete |
| Released | January 7, 2025 | January 7, 2025 |
| Launch price (MSRP) | $599 | $1999 |
| Case | Desktop | Desktop |
| Purpose | Gaming | Gaming |
| Segment | High-end | High-end |
| Architecture | RDNA 4.0 | Blackwell 2.0 |
| GPU Codename | Navi 48 | GB202-300 |
| Rival Equivalent | - GeForce RTX 5070 Ti | - |
| Recommended CPU | - Intel Core Ultra 5 245K or above | - Intel Core Ultra 9 285K or above |
Desktop GPU rating (12th and 2nd place)
Graphics Processing Unit
| Base Clock | 1660 MHz | 2017 MHz |
| Boost Clock | 2970 MHz | 2407 MHz |
| Shading Units | 4096 | 21760 |
| Texture Mapping Units (TMUs) | 256 | 680 |
| Render Output Units (ROPs) | 128 | 176 |
| Compute Units (Pipelines) | 64 | 170 |
| Tensor Cores | 128 | 680 |
| Ray-tracing Cores | 64 | 170 |
| L1 Cache | 128KB per cluster | 128KB per cluster |
| L2 Cache | 8MB shared | 96MB shared |
| Instructions Per Cycle | 4 IPC | 2 IPC |
Raw Performance
| Pixel Fill Rate | 380 GPixel/s | 424 GPixel/s |
| Texture Fill Rate | 760 GTexel/s | 1637 GTexel/s |
FLOPS (FP32)
Differences in GPUs' instructions per clock make TFLOPS values not directly comparable.
Physical
| Interface | PCIe 5.0 x16 | PCIe 5.0 x16 |
| TGP | 304 W | 575 W |
| Manufacturing | TSMC | TSMC |
| Fabrication Process | 4 nm | 4 nm |
| Die Size | 357 mm² | 750 mm² |
| Transistor Count | 53.9 billion | 92.2 billion |
| Transistor Density | 150.98 MTr/mm² | 122.93 MTr/mm² |
Memory
| Memory Type | GDDR6 | GDDR7 |
| Memory Size | 16 GB | 32 GB |
| Memory Clock | 2518 MHz | 1750 MHz |
| Effective Memory Speed | 20100 Mbps | 28000 Mbps |
| Bus | 256-bit | 512-bit |
| ECC | No | No |
Memory Bandwidth
API
| DirectX | 12.2 | 12.2 |
| Vulkan | 1.4 | 1.4 |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 2.2 | 3.0 |
| CUDA | No | 12.0 |
| Ray Tracing | Yes | Yes |
| DLSS | No | DLSS 4 |
| DisplayPort | 2.1a | 2.1b |
Cast your vote
Total votes: 236