- Graphics Processor: Navi 48
- Cores: 4096
- TMUs / ROPs: 256 / 128
Expert Verdict
A mid-to-upper-range AMD Radeon RX 9070 XT graphics card with capable gaming performance (73/100). Suited for 1440p/1080p gaming and content creation.
Review
Gaming
73
Workstation
62
AI/ML
67
Energy Efficiency
49
Quick Comparison Final Score
68
Pros
- Solid gaming performance
- Large 16GB memory pool
- Hardware ray tracing support
- Dedicated AI tensor cores
Cons
- High power consumption (304W TGP)
- No DLSS upscaling support
Compare Radeon RX 9070 XT with other products
Full Specifications
Games
Performance
Forza Horizon 5
The Witcher 3
Counter-Strike 2
Far Cry 6
Hogwarts Legacy
Call of Duty
Ghost of Tsushima
Cyberpunk 2077
Tomb Raider
Average FPS
1080p High
1080p Ultra
1440p Ultra
4K Ultra
FPS Table
| 1080p High | 1080p Ultra | 1440p Ultra | 4K Ultra | |
|---|---|---|---|---|
| Forza Horizon 5 | 169 | — | — | — |
| The Witcher 3 | 236 | — | — | — |
| Counter-Strike 2 | 224 | — | — | — |
| Far Cry 6 | 178 | — | — | — |
| Hogwarts Legacy | 132 | — | — | — |
| Call of Duty | 234 | — | — | — |
| Ghost of Tsushima | 135 | — | — | — |
| Cyberpunk 2077 | 166 | — | — | — |
| Tomb Raider | 268 | — | — | — |
| Average FPS | 194 | — | — | — |
Average FPS by Resolution
1080p High
229
1080p Ultra
194
1440p Ultra
150
4K Ultra
89
Margin of Error
Low
3D Mark
Multiplatform graphics benchmark suite that directly correlates with performance in modern games
| Steel Nomad Lite Score | 26180 |
|---|---|
| Time Spy | 29701 |
| Solar Bay | 121225 |
| Port Royal | 18525 |
| Fire Strike | 66381 |
| Wild Life Extreme | 54685 |
| Night Raid | 195698 |
GeekBench 6 OpenCL
GPU test for computational tasks (image processing, photography, computer vision, and ML)
| GB6 Compute Score | 182816 |
|---|---|
| Background Blur | 427.4 img/sec |
| Face Detection | 251 img/sec |
| Horizon Detection | 6.77 Gpixels/sec |
| Edge Detection | 9.42 Gpixels/sec |
| Gaussian Blur | 9.35 Gpixels/sec |
| Feature Matching | 1.42 Gpixels/sec |
| Stereo Matching | 651.4 Gpixels/sec |
| Particle Physics | 23630 FPS |
| API | OpenCL |
Passmark Graphics
Videocard test that focuses on compute shaders, multi-texturing, tessellation, and other features
Blender
Rendering performance test for 3D modeling
GeekBench 6 ML
Tests throughput of AI operations in single, half, and quantized precision
| GB6 ML Single Precision | 39190 |
|---|---|
| GB6 ML Half Precision | 55727 |
| GB6 ML Quantized | 33599 |
| Image Classification (SP) | 18001 |
| Image Segmentation (HP) | 25174 |
| Image Super Resolution (Q) | 45714 |
| Face Detection (HP) | 74776 |
| Pose Estimation (Q) | 196137 |
| Text Classification (SP) | 5163 |
| Machine Translation (HP) | 8677 |
| Object Detection (SP) | 19602 |
| Depth Estimation (Q) | 74879 |
| Style Transfer (SP) | 411302 |
| Framework | ONNX |
| Backend | DirectML |
General
| Vendor | Amd |
|---|---|
| Build | Discrete |
| Released | 7-Jan-25 |
| Case | Desktop |
| Purpose | Gaming |
| Segment | High-end |
| Architecture | RDNA 4.0 |
| GPU Codename | Navi 48 |
| Recommended CPU | -Intel Core Ultra 5 245Kor above |
Graphics Processing Unit
| Base Clock | 1660 MHz |
|---|---|
| Boost Clock | 2970 MHz |
| Shading Units | 4096 |
| Texture Mapping Units (TMUs) | 256 |
| Render Output Units (ROPs) | 128 |
| Compute Units (Pipelines) | 64 |
| Tensor Cores | 128 |
| Ray-tracing Cores | 64 |
| L1 Cache | 128KB per cluster |
| L2 Cache | 4MB shared |
| Instructions Per Cycle | 4 IPC |
Raw Performance
| Pixel Fill Rate | 380 GPixel/s |
|---|---|
| Texture Fill Rate | 760 GTexel/s |
| FLOPS (FP32) | 48.7 TFLOPS |
Physical
| Interface | PCIe 5.0 x16 |
|---|---|
| TGP | 304 W |
| Manufacturing | TSMC |
| Fabrication Process | 4 nm |
| Die Size | 356.5 mm² |
| Transistor Count | 53.9 billion |
| Transistor Density | 151.19 MTr/mm² |
| Max. Temperature | — |
Memory
| Memory Type | GDDR6 |
|---|
API
| DirectX | 12 |
|---|---|
| Vulkan | 1.3 |
| OpenGL | 4.6 |
| OpenCL | 2.2 |
| CUDA | No |
| Ray Tracing | Yes |
| DLSS | No |
| DisplayPort | 2.1a |
