GeForce RTX 4070 Ti SUPER vs RTX 3090
VS
We compared two discrete desktop gaming GPUs: the GeForce RTX 4070 Ti SUPER 16 GB with 66 pipelines and 8448 shaders against the 3 years and 5 months older RTX RTX 3090 24 GB that utilizes 82 pipelines and 10496 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 4070 Ti SUPER
- Performs slightly better (up to 18%) in 3DMark Steel Nomad Lite
- Manufactured using a more efficient 5 nm process technology
- Shows 12% higher average frame rate in modern games at QHD resolution – 134 vs 120 FPS
- 24% higher maximum theoretical performance (44.1 vs 35.6 TFLOPS)
- Supports Nvidia DLSS 3 technology
Reasons to consider the GeForce RTX 3090
- Includes 8 GB more video memory
- Has 39% higher memory bandwidth: 936.2 vs 672.3 GB/s
Gaming Performance
Frame rate comparison across popular AAA titles at different resolutionsGames
| Forza Horizon 5 | 210 148 140 107 |
186 128 117 91 |
| The Witcher 3 | 221 192 140 72 |
205 170 128 66 |
| Counter-Strike 2 | 294 219 164 83 |
275 207 155 83 |
| Far Cry 6 | 168 156 138 83 |
156 145 124 77 |
| Hogwarts Legacy | 146 122 94 54 |
128 108 86 47 |
| CoD: Modern Warfare III | 195 189 144 99 |
171 159 123 80 |
| Ghost of Tsushima | 134 113 95 62 |
120 97 85 48 |
| Cyberpunk 2077 | 169 153 105 49 |
148 132 90 43 |
| Shadow of the Tomb Raider | 270 250 190 106 |
237 226 168 91 |
| 1080p High | 201 | 181 |
| 1080p Ultra | 171 | 152 |
| 1440p Ultra | 134 | 120 |
| 4K Ultra | 79 | 70 |
| Margin of Error | Low | Low |
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Time Spy | 24220 | 19896 |
| Solar Bay | 116588 | 97395 |
| Port Royal | 15806 | 13633 |
| Fire Strike | 56343 | 47512 |
| Wild Life Extreme | 49911 | 43669 |
| Night Raid | 176672 | 139766 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 331.7 img/sec | 320.8 img/sec |
| Face Detection | 211.6 img/sec | 207.2 img/sec |
| Horizon Detection | 9.63 Gpixels/sec | 10.1 Gpixels/sec |
| Edge Detection | 13.7 Gpixels/sec | 14.4 Gpixels/sec |
| Gaussian Blur | 13.9 Gpixels/sec | 11.5 Gpixels/sec |
| Feature Matching | 2.15 Gpixels/sec | 1.48 Gpixels/sec |
| Stereo Matching | 1210 Gpixels/sec | 984.2 Gpixels/sec |
| Particle Physics | 33566 FPS | 27764.9 FPS |
| API | OpenCL | OpenCL |
Cinebench 2024 GPU
Cinebench 2024 GPU
Passmark Graphics
G3D Mark Score
| G2D Mark | 1246 | 1069 |
| DirectX 11 | 277 FPS | 218 FPS |
| DirectX 12 | 119 FPS | 109 FPS |
| GPU Compute | 18206 Ops/s | 15028 Ops/s |
Blender
Blender GPU
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) | 12970 | 10665 |
| Image Segmentation (HP) | 26364 | 28963 |
| Image Super Resolution (Q) | 37217 | 22697 |
| Face Detection (HP) | 58720 | 55232 |
| Pose Estimation (Q) | 177604 | 128155 |
| Text Classification (SP) | 3815 | 3152 |
| Machine Translation (HP) | 5829 | 5088 |
| Object Detection (SP) | 15675 | 13891 |
| Depth Estimation (Q) | 62447 | 37984 |
| Style Transfer (SP) | 357918 | 272000 |
| Framework | ONNX | ONNX |
| Backend | DirectML | DirectML |
Specifications
Technical specifications of GeForce RTX 4070 Ti SUPER and RTX 3090General
| Vendor | Nvidia | Nvidia |
| Build | Discrete | Discrete |
| Released | January 24, 2024 | September 24, 2020 |
| Launch price (MSRP) | $799 | $1499 |
| Case | Desktop | Desktop |
| Purpose | Gaming | Gaming |
| Segment | High-end | High-end |
| Architecture | Ada Lovelace | Ampere |
| GPU Codename | AD103 | GA102 |
| Rival Equivalent | - Radeon RX 7900 XTX | - Radeon RX 6900 XT |
| Successor | - | - GeForce RTX 5090 |
| Recommended CPU | - Intel Core i7 14700K or above | - Intel Core i9 12900K or above |
Desktop GPU rating (13th and 20th place)
Graphics Processing Unit
| Base Clock | 2340 MHz | 1395 MHz |
| Boost Clock | 2610 MHz | 1695 MHz |
| Shading Units | 8448 | 10496 |
| Texture Mapping Units (TMUs) | 264 | 328 |
| Render Output Units (ROPs) | 96 | 112 |
| Compute Units (Pipelines) | 66 | 82 |
| Tensor Cores | 264 | 328 |
| Ray-tracing Cores | 66 | 82 |
| L1 Cache | 128KB per cluster | 128KB per cluster |
| L2 Cache | 48MB shared | 6MB shared |
| Instructions Per Cycle | 2 IPC | 2 IPC |
Raw Performance
| Pixel Fill Rate | 251 GPixel/s | 190 GPixel/s |
| Texture Fill Rate | 689 GTexel/s | 556 GTexel/s |
FLOPS (FP32)
Physical
| Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| TGP | 285 W | 350 W |
| Manufacturing | TSMC | Samsung |
| Fabrication Process | 5 nm | 8 nm |
| Die Size | 379 mm² | 628 mm² |
| Transistor Count | 45 billion | 28 billion |
| Transistor Density | 118.73 MTr/mm² | 44.59 MTr/mm² |
Memory
| Memory Type | GDDR6X | GDDR6X |
| Memory Size | 16 GB | 24 GB |
| Memory Clock | 1313 MHz | 1219 MHz |
| Effective Memory Speed | 21000 Mbps | 19500 Mbps |
| Bus | 256-bit | 384-bit |
| ECC | No | No |
Memory Bandwidth
API
| DirectX | 12 | 12 |
| Vulkan | 1.3 | 1.3 |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| CUDA | 8.9 | 8.6 |
| Ray Tracing | Yes | Yes |
| DLSS | DLSS 3 | DLSS 2 |
| DisplayPort | 1.4a | 1.4a |
Cast your vote
Total votes: 64
