Product Compare
NVIDIA Tesla P4 vs NVIDIA GeForce GTX 1650
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Sep 2016 | Apr 2019 |
| Generation | Tesla | GeForce 16 |
| Type | Professional | Desktop |
| Bus Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
| Clock Speeds | ||
| Base Clock | 886 MHz | 1485 MHz |
| Boost Clock | 1114 MHz | 1665 MHz |
| Memory Clock | 1502 MHz | 2001 MHz |
| Memory | ||
| Memory Size | 8GB | 4GB |
| Memory Type | GDDR5 | GDDR5 |
| Memory Bus | 256bit | 128bit |
| Bandwidth | 192.3GB/s | 128.1GB/s |
| Render Config | ||
| SM Count | 20 | 14 |
| Shading Units | 2560 | 896 |
| TMUs | 160 | 56 |
| ROPs | 64 | 32 |
| L1 Cache | 48 KB (per SM) | 64 KB (per SM) |
| L2 Cache | 2 MB | 1024 KB |
| Theoretical Performance | ||
| Pixel Rate | 71.30 GPixel/s | 53.28 GPixel/s |
| Texture Rate | 178.2 GTexel/s | 93.24 GTexel/s |
| FP16 (half) | 89.12 GFLOPS | 5.967 TFLOPS |
| FP32 (float) | 5.704 TFLOPS | 2.984 TFLOPS |
| FP64 (double) | 178.2 GFLOPS | 93.24 GFLOPS |
| Graphics Processor | ||
| GPU Name | GP104 | TU117 |
| GPU Variant | GP104-895-A1 | TU117-300-A1 |
| Architecture | Pascal | Turing |
| Foundry | TSMC | TSMC |
| Process Size | 16 nm | 12 nm |
| Transistors | 7.2 billion | 4.7 billion |
| Die Size | 314 mm² | 200 mm² |
| Board Design | ||
| TDP | 75W | 75W |
| Suggested PSU | 250 W | 250 W |
| Outputs | No outputs | 1x DVI1x HDMI 2.01x DisplayPort 1.4a |
| Power Connectors | None | None |
| Graphics Features | ||
| DirectX | 12 (12_1) | 12 (12_1) |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.3 | 1.3 |
| CUDA | 6.1 | 7.5 |
| Shader Model | 6.4 | 6.6 |