Quick/Comparisons
The GeForce RTX 3070 Laptop's superior gaming score is largely due to its Ampere architecture and 8GB GDDR6 memory, which provide a significant boost in frame rates and graphics quality.
The GeForce RTX 3070 Laptop's higher workstation score is attributed to its GA104 (GN20-E5) codename and PCIe 4.0 x16 interface, which enable faster data transfer and processing.
The GeForce RTX 3070 Laptop's AI/ML score is higher due to its dedicated AI tensor cores and support for hardware ray tracing, which enhance its capabilities in AI and machine learning tasks.
34 | Gaming | 14 |
41 | Workstation | 24 |
31 | AI/ML |
Compare with something else?
Compare GeForce RTX 3070 with another →14 |
43 | Energy Efficiency | 24 |
37 | Quick Comparison Final Score | 17 |
| GA104 (GN20-E5) | Graphics Processor | Polaris 20 |
| 5120 | Cores | 2304 |
| 160 / 80 | TMUs / ROPs | 144 / 32 |
| 2998.92 | Blender GPU | 391.86 |
| 79 | Forza Horizon 5 | 38 |
| 114 | The Witcher 3 | 41 |
| 150 | Counter-Strike 2 | 56 |
| 94 | Far Cry 6 | 46 |
| 61 | Hogwarts Legacy | 28 |
| 88 | Call of Duty | 41 |
| 56 | Ghost of Tsushima | 24 |
| 73 | Cyberpunk 2077 | 28 |
| 116 | Tomb Raider | 52 |
| 92 | Average FPS | 39 |
| 114 | 1080p High | 51 |
| 92 | 1080p Ultra | 39 |
| 68 | 1440p Ultra | 26 |
| 38 | 4K Ultra | 13 |
| Low | Margin of Error | Low |
| 118499 | GB6 Compute Score | 40514 |
| 170.3 img/sec | Background Blur | 59.5 img/sec |
| 113.6 img/sec | Face Detection | 33.8 img/sec |
| 5.59 Gpixels/sec | Horizon Detection | 1.56 Gpixels/sec |
| 7.94 Gpixels/sec | Edge Detection | 3.1 Gpixels/sec |
| 6.19 Gpixels/sec | Gaussian Blur | 2.92 Gpixels/sec |
| 1.16 Gpixels/sec | Feature Matching | 0.33 Gpixels/sec |
| 467 Gpixels/sec | Stereo Matching | 161.8 Gpixels/sec |
| 15131.3 FPS | Particle Physics | 5395.1 FPS |
| OpenCL | API | OpenCL |
| 15105 | G3D Mark Score | 8792 |
| 630 | G2D Mark | 769 |
| 137 FPS | DirectX 11 | 60 FPS |
| 63 FPS | DirectX 12 | 43 FPS |
| 6702 Ops/s | GPU Compute | 3431 Ops/s |
| 17250 | GB6 ML Single Precision | 7905 |
| 27936 | GB6 ML Half Precision | 8166 |
| 11249 | GB6 ML Quantized | 5642 |
| 6969 | Image Classification (SP) | 3143 |
| 17142 | Image Segmentation (HP) | 3955 |
| 16426 | Image Super Resolution (Q) | 7872 |
| 31628 | Face Detection (HP) | 11922 |
| 74413 | Pose Estimation (Q) | 41162 |
| 2522 | Text Classification (SP) | 1199 |
| 3639 | Machine Translation (HP) | 1773 |
| 9002 | Object Detection (SP) | 3620 |
| 27790 | Depth Estimation (Q) | 13995 |
| 160005 | Style Transfer (SP) | 84083 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 103-130 GPixel/s | Pixel Fill Rate | 43 GPixel/s |
| 206-259 GTexel/s | Texture Fill Rate | 193 GTexel/s |
| 13.2-16.6 TFLOPS | FLOPS (FP32) | 6.2 TFLOPS |
| GDDR6 | Memory Type | GDDR5 |
| 8 GB | Memory Size | 8 GB |
| 1500-1750 MHz | Memory Clock | 2000 MHz |
| 12000-14000 Mbps | Effective Memory Speed | 8000 Mbps |
| 256-bit | Bus | 256-bit |
| No | ECC | No |
| 448 GB/s | Memory Bandwidth | 256 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 3.0 x16 |
| 80-150 W (configurable) | TGP | 185 W |
| Samsung | Manufacturing | GlobalFoundries |
| 8 nm | Fabrication Process | 14 nm |
| 392 mm² | Die Size | 232 mm² |
| 17.4 billion | Transistor Count | 5 billion |
| 44.39 MTr/mm² | Transistor Density | 21.55 MTr/mm² |
| 12 | DirectX | 12 |
| 1.2 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 2 | OpenCL | 2.1 |
| 8.6 | CUDA | — |
| Yes | Ray Tracing | No |
| DLSS 2 | DLSS | No |
| 1.4a | DisplayPort | 1.4a |
| Nvidia | Vendor | Amd |
| Discrete | Build | Discrete |
| January 12, 2021 | Released | April 18, 2017 |
| Laptop | Case | Desktop |
| Gaming | Purpose | Gaming |
| High-end | Segment | High-end |
| Ampere | Architecture | GCN 4.0 |
| GA104 (GN20-E5) | GPU Codename | Polaris 20 |
| -Intel Core i7 12700Hor above | Recommended CPU | -Intel Core i5 8500or above |
| 10214 | Steel Nomad Lite Score | 4182 |
| 10477 | Time Spy | 4354 |
| 45215 | Solar Bay | — |
| 6247 | Port Royal | 527 |
| 26729 | Fire Strike | 14979 |
| 21137 | Wild Life Extreme | 8920 |
| 82533 | Night Raid | 56872 |
| 780-1215 MHz (configurable) | Base Clock | 1257 MHz |
| 1290-1620 MHz (configurable) | Boost Clock | 1340 MHz |
| 5120 | Shading Units | 2304 |
| 160 | Texture Mapping Units (TMUs) | 144 |
| 80 | Render Output Units (ROPs) | 32 |
| 40 | Compute Units (Pipelines) | — |
| 160 | Tensor Cores | No |
| 40 | Ray-tracing Cores | No |
| 128KB per cluster | L1 Cache | 16KB per cluster |
| 4MB shared | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |