At the end of last month, the 60-series GeForce RTX finally came out, which has had high popularity every generation. The RTX 3060 is the cheapest Ampere graphics card so far, and the Gigabyte Eagle (OC) version is also relatively power-efficient. Both in terms of smaller dimensions and relatively lower power draw. However, a detailed analysis is required for the bigger picture, to which we’re inviting you.
| Parameters | Gigabyte RTX 3060 Eagle OC 12G | |
| Architecture | Ampere | |
| Die | GA106 | |
| Manufacturing node | 8 nm Samsung | |
| Die size | 276 mm² | |
| Transistor count | 13,25 mld. | |
| Compute units | 28 SM | |
| Shaders/CUDA cores | 3584 | |
| Base Clock | 1320 MHz | |
| Game Clock (AMD) | – | |
| Boost Clock | 1807 MHz | |
| RT units | 28 | |
| AI/tensor cores | 112 | |
| ROPs | 48 | |
| TMUs | 112 | |
| L2 Cache | 3 MB | |
| Infinity Cache | – | |
| Interface | PCIe 4.0 ×16 | |
| Multi-GPU interconnect | – | |
| Memory | 12 GB GDDR6 | |
| Memory clock (effective) | 15 GHz | |
| Memory bus | 192 bitov | |
| Memory bandwidth | 360 GB/s | |
| Pixel fillrate | 86,7 Gpx/s | |
| Texture fillrate | 202,4 Gtx/s | |
| FLOPS (FP32) | 12,74 TFLOPS | |
| FLOPS (FP64) | 199,0 GFLOPS | |
| FLOPS (FP16) | 12,74 TFLOPS | |
| AI/tensor TOPS (INT8) | 101,92 TOPS | |
| AI/tensor FLOPS (FP16) | 50,96 TFLOPS | |
| TDP | 170 W | |
| Power connectors | 1× 8pin | |
| Card lenght | 242 mm | |
| Card slots used | 41 mm | |
| Shader Model | 6.5 | |
| DirectX/Feature Level | DX 12 Ultimate (12_2) | |
| OpenGL | 4.6 | |
| Vulkan | 1.2 | |
| OpenCL | 2.0 | |
| CUDA | 8.6 | |
| Video encoder engine | NVEnc 7 | |
| Encoding formats | HEVC, H.264 | |
| Encoding resolution | 8K | |
| Video decoder engine | NVDec 5 | |
| Decoding formats | HEVC,H.264,VP9, AV1 | |
| Decoding resolution | 8K | |
| Max. Monitor resolution | 7680 × 4320 px | |
| HDMI | 1.4a (2×) | |
| DisplayPort | 2.1 (2×) | |
| USB-C | – | |
| MSRP | 443 euros |
