NVIDIA GeForce RTX 3060 Ti vs NVIDIA GeForce RTX 3070 Mobile
What is the difference between NVIDIA GeForce RTX 3060 Ti and NVIDIA GeForce RTX 3070 Mobile. Find out which graphics card has better performance.
Benchmarks
| 19650 | PassMark G3D | 14483 |
| 936 | PassMark G2D | 627 |
Graphics Processor (GPU)
| GA104 | GPU Name | GA104 |
| GA104-200-A1 | Model number | GA104-770-A1 |
| Ampere | Architecture | Ampere |
| Samsung | Foundry | Samsung |
| 8 nm | Process Size | 8 nm |
| 17,400 million | Transistors | 17,400 million |
| 392 mm² | Die Size | 392 mm² |
Graphics Card
| Dec 1st, 2020 | Release Date | Jan 12th, 2021 |
| GeForce 30 | Family | GeForce 30 Mobile |
| Active | Production | Active |
| Dec 2nd, 2020 | Availability | Feb 2nd, 2021 |
| GeForce 20 | Predecessor | GeForce 20 Mobile |
| PCIe 4.0 x16 | Bus Interface | PCIe 4.0 x16 |
Memory
| 8 GB | Memory Size | 8 GB |
| GDDR6 | Memory Type | GDDR6 |
| 256 bit | Memory Bus | 256 bit |
| 448.0 GB/s | Bandwidth | 448.0 GB/s |
Performance
| 133.2 GPixel/s | Pixel fillrate | 124.8 GPixel/s |
| 253.1 GTexel/s | Texture fillrate | 249.6 GTexel/s |
| 16.20 TFLOPS (1:1) | FP16 (half) performance | 15.97 TFLOPS (1:1) |
| 16.20 TFLOPS | FP32 (float) performance | 15.97 TFLOPS |
| 253.1 GFLOPS (1:64) | FP64 (double) performance | 249.6 GFLOPS (1:64) |
Clock Speeds
| 1410 MHz | Base Clock | 1110 MHz |
| 1665 MHz | Boost Clock | 1560 MHz |
| 1750 MHz 14 Gbps effective | Memory Clock | 1750 MHz 14 Gbps effective |
Render Config
| 4864 | Shading Units | 5120 |
| 152 | TMUs | 160 |
| 80 | ROPs | 80 |
| 38 | RT Cores | 40 |
| 128 KB (per SM) | L1 Cache | 128 KB (per SM) |
| 4 MB | L2 Cache | 4 MB |
| 38 | SM Count | 40 |
| 152 | Tensor Cores | 160 |
Board Design
| 200 W | Thermal design power (TDP) | 115 W |
| 1x HDMI 3x DisplayPort | Display Connectors | No outputs |
| 1x 12-pin | Power Connectors | None |
API support
| 12 Ultimate (12_2) | DirectX | 12 Ultimate (12_2) |
| 4.6 | OpenGL | 4.6 |
| 3.0 | OpenCL | 3.0 |
| 1.2 | Vulkan | 1.2 |
| 6.6 | Shader Model | 6.6 |
| 8.6 | CUDA | 8.6 |