Product Compare
NVIDIA H200 NVL vs NVIDIA GeForce RTX 3060 3840SP
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Nov 2024 | Jan 2021 |
| Generation | Server Hopper | GeForce 30 |
| Type | AI GPU | Desktop |
| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 1365 MHz | 1627 MHz |
| Boost Clock | 1785 MHz | 1852 MHz |
| Memory Clock | 1593 MHz | 1750 MHz |
| Memory | ||
| Memory Size | 141GB | 6GB |
| Memory Type | HBM3e | GDDR6 |
| Memory Bus | 6144bit | 192bit |
| Bandwidth | 4.89TB/s | 336.0GB/s |
| Render Config | ||
| SM Count | 132 | 30 |
| Shading Units | 16896 | 3840 |
| TMUs | 528 | 120 |
| ROPs | 24 | 48 |
| Tensor Cores | 528 | 120 |
| L1 Cache | 256 KB (per SM) | 128 KB (per SM) |
| L2 Cache | 50 MB | 3 MB |
| RT Cores | 30 | |
| Theoretical Performance | ||
| Pixel Rate | 42.84 GPixel/s | 88.90 GPixel/s |
| Texture Rate | 942.5 GTexel/s | 222.2 GTexel/s |
| FP16 (half) | 241.3 TFLOPS | 14.22 TFLOPS |
| FP32 (float) | 60.32 TFLOPS | 14.22 TFLOPS |
| FP64 (double) | 30.16 TFLOPS | 222.2 GFLOPS |
| Graphics Processor | ||
| GPU Name | GH100 | GA106 |
| Architecture | Hopper | Ampere |
| Foundry | TSMC | Samsung |
| Process Size | 5 nm | 8 nm |
| Transistors | 80 billion | 12 billion |
| Die Size | 814 mm² | 276 mm² |
| GPU Variant | GA106-400-A1 | |
| Board Design | ||
| TDP | 600W | 185W |
| Suggested PSU | 1000 W | 450 W |
| Outputs | No outputs | 1x HDMI 2.13x DisplayPort 1.4a |
| Power Connectors | 8-pin EPS | 1x 12-pin |
| Graphics Features | ||
| DirectX | N/A | 12 Ultimate (12_2) |
| OpenGL | N/A | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | N/A | 1.3 |
| CUDA | 9.0 | 8.6 |
| Shader Model | N/A | 6.7 |