GeForce RTX 4070 Ti vs 3090 Ti
VS
We compared two discrete desktop gaming GPUs: the GeForce RTX 4070 Ti 12 GB with 60 pipelines and 7680 shaders against the 10 months older RTX 3090 Ti 24 GB that utilizes 84 pipelines and 10752 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
- Manufactured using a more efficient 5 nm process technology
- Supports Nvidia DLSS 3 technology
Reasons to consider the GeForce RTX 3090 Ti
- Includes 12 GB more video memory
- Has 2x higher memory bandwidth: 1008 vs 504 GB/s
- Features 96 more tensor cores for effective ML and AI workloads
- Has 40% more shading units (10752 vs 7680)
Gaming Performance
Frame rate comparison across popular AAA titles at different resolutionsGames
| Forza Horizon 5 | 197 136 126 100 |
200 136 126 99 |
| The Witcher 3 | 211 190 135 72 |
214 185 130 71 |
| Counter-Strike 2 | 288 218 160 83 |
286 213 154 85 |
| Far Cry 6 | 164 147 130 81 |
160 151 133 83 |
| Hogwarts Legacy | 140 115 89 50 |
141 116 93 52 |
| CoD: Modern Warfare III | 181 172 134 90 |
186 174 139 90 |
| Ghost of Tsushima | 124 107 92 55 |
129 105 89 57 |
| Cyberpunk 2077 | 161 147 98 44 |
158 140 98 44 |
| Shadow of the Tomb Raider | 257 243 188 98 |
254 237 183 101 |
| 1080p High | 191 | 192 |
| 1080p Ultra | 164 | 162 |
| 1440p Ultra | 128 | 127 |
| 4K Ultra | 75 | 76 |
| Margin of Error | Low | Low |
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Time Spy | 22743 | 21855 |
| Solar Bay | 108437 | 109239 |
| Port Royal | 13995 | 14890 |
| Fire Strike | 53473 | 51935 |
| Wild Life Extreme | 45396 | 48861 |
| Night Raid | 171208 | 156537 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 371.2 img/sec | 362.7 img/sec |
| Face Detection | 235 img/sec | 227.9 img/sec |
| Horizon Detection | 7.51 Gpixels/sec | 11.3 Gpixels/sec |
| Edge Detection | 10.6 Gpixels/sec | 15.9 Gpixels/sec |
| Gaussian Blur | 12.4 Gpixels/sec | 12.1 Gpixels/sec |
| Feature Matching | 2.03 Gpixels/sec | 1.56 Gpixels/sec |
| Stereo Matching | 1120 Gpixels/sec | 1080 Gpixels/sec |
| Particle Physics | 30420.2 FPS | 31024.6 FPS |
| API | OpenCL | OpenCL |
Cinebench 2024 GPU
Cinebench 2024 GPU
Passmark Graphics
G3D Mark Score
| G2D Mark | 1210 | 1226 |
| DirectX 11 | 288 FPS | 238 FPS |
| DirectX 12 | 115 FPS | 121 FPS |
| GPU Compute | 18158 Ops/s | 16494 Ops/s |
Blender
Blender GPU
Recent User Tests
GeForce RTX 4070 Ti
| Date | Benchmark | Result |
|---|---|---|
| 📘 2026-09-14 (Isa) | Cinebench 2024 | 80717 |
| 📘 2026-06-02 (Harley) | Cinebench 2024 | 79993 |
| 📘 2026-03-27 (minato) | Cinebench 2024 | 80470 |
GeForce RTX 3090 Ti
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) | 12496 | 10246 |
| Image Segmentation (HP) | 33378 | 22830 |
| Image Super Resolution (Q) | 37180 | 23532 |
| Face Detection (HP) | 62919 | 48657 |
| Pose Estimation (Q) | 171125 | 147241 |
| Text Classification (SP) | 3630 | 2917 |
| Machine Translation (HP) | 6420 | 4427 |
| Object Detection (SP) | 15804 | 13323 |
| Depth Estimation (Q) | 58647 | 40055 |
| Style Transfer (SP) | 331379 | 298683 |
| Framework | ONNX | ONNX |
| Backend | DirectML | DirectML |
Specifications
Technical specifications of GeForce RTX 4070 Ti and 3090 TiGeneral
| Vendor | Nvidia | Nvidia |
| Build | Discrete | Discrete |
| Released | January 3, 2023 | March 29, 2022 |
| Launch price (MSRP) | $799 | $1999 |
| Case | Desktop | Desktop |
| Purpose | Gaming | Gaming |
| Segment | Mid-range | High-end |
| Architecture | Ada Lovelace | Ampere |
| GPU Codename | AD104 | GA102 |
| Rival Equivalent | - Radeon RX 7900 XT | - |
| Successor | - GeForce RTX 5070 Ti | - |
| Recommended CPU | - Intel Core i7 14700K or above | - Intel Core i9 12900K or above |
Desktop GPU rating (15th and 14th place)
Graphics Processing Unit
| Base Clock | 2310 MHz | 1560 MHz |
| Boost Clock | 2610 MHz | 1860 MHz |
| Shading Units | 7680 | 10752 |
| Texture Mapping Units (TMUs) | 240 | 336 |
| Render Output Units (ROPs) | 80 | 112 |
| Compute Units (Pipelines) | 60 | 84 |
| Tensor Cores | 240 | 336 |
| Ray-tracing Cores | 60 | 84 |
| 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 | 209 GPixel/s | 208 GPixel/s |
| Texture Fill Rate | 626 GTexel/s | 625 GTexel/s |
FLOPS (FP32)
Physical
| Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| TGP | 285 W | 450 W |
| Manufacturing | TSMC | Samsung |
| Fabrication Process | 5 nm | 8 nm |
| Die Size | 294 mm² | 628 mm² |
| Transistor Count | 35 billion | 28 billion |
| Transistor Density | 119.05 MTr/mm² | 44.59 MTr/mm² |
Memory
| Memory Type | GDDR6X | GDDR6X |
| Memory Size | 12 GB | 24 GB |
| Memory Clock | 1313 MHz | 1313 MHz |
| Effective Memory Speed | 21000 Mbps | 21000 Mbps |
| Bus | 192-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: 69