AMD Radeon RX 9070 vs GeForce RTX 4070
VS
AMD Radeon RX 9070
Navi 48
Nvidia GeForce RTX 4070
AD104
We compared two discrete desktop gaming GPUs: the Radeon RX 9070 16 GB with 56 pipelines and 3584 shaders against the 1 year and 9 months older GeForce RTX 4070 12 GB that utilizes 46 pipelines and 5888 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 Radeon RX 9070
- Performs better (up to 26%) in 3DMark Steel Nomad Lite
- Shows 31% higher average frame rate in modern games at QHD resolution – 139 vs 106 FPS
- Manufactured using a more efficient 4 nm process technology
- 24% higher maximum theoretical performance (36.1 vs 29.1 TFLOPS)
- Includes 4 GB more video memory
- Has 28% higher memory bandwidth: 644.6 vs 504 GB/s
Reasons to consider the GeForce RTX 4070
- Supports Nvidia DLSS 3 technology
- Features 72 more tensor cores for effective ML and AI workloads
- Has 64% more shading units (5888 vs 3584)
Gaming Performance
Frame rate comparison across popular AAA titles at different resolutionsGames
| Forza Horizon 5 | 229 156 137 102 |
172 115 102 79 |
| The Witcher 3 | 240 216 159 84 |
194 159 114 61 |
| Counter-Strike 2 | 307 215 158 84 |
272 205 149 81 |
| Far Cry 6 | 192 163 148 95 |
149 133 115 68 |
| Hogwarts Legacy | 146 119 97 52 |
117 97 77 39 |
| CoD: Modern Warfare III | 219 212 157 107 |
149 145 107 69 |
| Ghost of Tsushima | 146 122 105 65 |
106 88 72 43 |
| Cyberpunk 2077 | 174 149 102 47 |
129 119 75 37 |
| Shadow of the Tomb Raider | 262 248 188 99 |
211 190 145 80 |
| 1080p High | 213 | 167 |
| 1080p Ultra | 178 | 139 |
| 1440p Ultra | 139 | 106 |
| 4K Ultra | 82 | 62 |
| Margin of Error | Low | Low |
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Time Spy | 26693 | 17799 |
| Solar Bay | 106118 | 82861 |
| Port Royal | 15939 | 11156 |
| Fire Strike | 60053 | 44067 |
| Wild Life Extreme | 48323 | 35472 |
| Night Raid | 177554 | 153385 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 377.5 img/sec | 233 img/sec |
| Face Detection | 221.3 img/sec | 141.7 img/sec |
| Horizon Detection | 6.4 Gpixels/sec | 6.62 Gpixels/sec |
| Edge Detection | 8.99 Gpixels/sec | 9.19 Gpixels/sec |
| Gaussian Blur | 8.52 Gpixels/sec | 9.2 Gpixels/sec |
| Feature Matching | 1.36 Gpixels/sec | 1.78 Gpixels/sec |
| Stereo Matching | 606 Gpixels/sec | 845.4 Gpixels/sec |
| Particle Physics | 8521.2 FPS | 25043.9 FPS |
| API | OpenCL | OpenCL |
Cinebench 2024 GPU
Cinebench 2024 GPU
Passmark Graphics
G3D Mark Score
| G2D Mark | 1262 | 1165 |
| DirectX 11 | 280 FPS | 243 FPS |
| DirectX 12 | 73 FPS | 103 FPS |
| GPU Compute | 14799 Ops/s | 14561 Ops/s |
Blender
Blender GPU
Recent User Tests
Radeon RX 9070
No benchmark results yet
GeForce RTX 4070
| Date | Benchmark | Result |
|---|---|---|
| 📘 2025-12-14 (kashitsu) | Cinebench 2024 | 18051 |
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) | 15375 | 10246 |
| Image Segmentation (HP) | 37834 | 26153 |
| Image Super Resolution (Q) | 43910 | 32054 |
| Face Detection (HP) | 67858 | 51337 |
| Pose Estimation (Q) | 172939 | 124617 |
| Text Classification (SP) | 4469 | 3703 |
| Machine Translation (HP) | 8499 | 5484 |
| Object Detection (SP) | 15393 | 13134 |
| Depth Estimation (Q) | 66890 | 50548 |
| Style Transfer (SP) | 365633 | 263477 |
| Framework | ONNX | ONNX |
| Backend | DirectML | DirectML |
Specifications
Technical specifications of Radeon RX 9070 and GeForce RTX 4070General
| Vendor | Amd | Nvidia |
| Build | Discrete | Discrete |
| Released | January 7, 2025 | April 13, 2023 |
| Launch price (MSRP) | $549 | $599 |
| Case | Desktop | Desktop |
| Purpose | Gaming | Gaming |
| Segment | High-end | Mid-range |
| Architecture | RDNA 4.0 | Ada Lovelace |
| GPU Codename | Navi 48 | AD104 |
| Rival Equivalent | - GeForce RTX 5070 | - Radeon RX 7900 XT |
| Successor | - | - GeForce RTX 5070 |
| Recommended CPU | - Intel Core Ultra 5 245K or above | - Intel Core i7 14700K or above |
Desktop GPU rating (16th and 30th place)
Graphics Processing Unit
| Base Clock | 1330 MHz | 1920 MHz |
| Boost Clock | 2520 MHz | 2475 MHz |
| Shading Units | 3584 | 5888 |
| Texture Mapping Units (TMUs) | 224 | 184 |
| Render Output Units (ROPs) | 128 | 64 |
| Compute Units (Pipelines) | 56 | 46 |
| Tensor Cores | 112 | 184 |
| Ray-tracing Cores | 56 | 46 |
| L1 Cache | 128KB per cluster | 128KB per cluster |
| L2 Cache | 8MB shared | 36MB shared |
| Instructions Per Cycle | 4 IPC | 2 IPC |
Raw Performance
| Pixel Fill Rate | 323 GPixel/s | 158 GPixel/s |
| Texture Fill Rate | 564 GTexel/s | 455 GTexel/s |
FLOPS (FP32)
Differences in GPUs' instructions per clock make TFLOPS values not directly comparable.
Physical
| Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| TGP | 220 W | 200 W |
| Manufacturing | TSMC | TSMC |
| Fabrication Process | 4 nm | 5 nm |
| Die Size | 357 mm² | 294 mm² |
| Transistor Count | 53.9 billion | 35 billion |
| Transistor Density | 150.98 MTr/mm² | 119.05 MTr/mm² |
Memory
| Memory Type | GDDR6 | GDDR6X |
| Memory Size | 16 GB | 12 GB |
| Memory Clock | 2518 MHz | 1313 MHz |
| Effective Memory Speed | 20100 Mbps | 21000 Mbps |
| Bus | 256-bit | 192-bit |
| ECC | No | No |
Memory Bandwidth
API
| DirectX | 12.2 | 12 |
| Vulkan | 1.4 | 1.3 |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 2.2 | 3.0 |
| CUDA | No | 8.9 |
| Ray Tracing | Yes | Yes |
| DLSS | No | DLSS 3 |
| DisplayPort | 2.1a | 1.4a |
Cast your vote
Total votes: 110
