NVIDIA P104-100 vs NVIDIA P106-100
Graphics Processor
| GPU model | GP104 | GP106 |
|---|---|---|
| GPU variant | GP104-100-A1 | GP106-100-A1 |
| Architecture | Pascal | Pascal |
| Foundry | TSMC | TSMC |
| Manufacturing process | 16 nm | 16 nm |
| Number of transistors | 7,200 million | 4,400 million |
| Die size | 314 mm² | 200 mm² |
Graphics Card
| Release date | Dec 12th, 2017 | Jun 19th, 2017 |
|---|---|---|
| Generation | Mining GPUs | Mining GPUs |
| Production | End-of-life | End-of-life |
| Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Clocks
| Base clock | 1607 MHz | 1506 MHz |
|---|---|---|
| Boost clock | 1733 MHz | 1709 MHz |
| Memory clock | 1251 MHz 10 Gbps effective | 2002 MHz 8 Gbps effective |
Memory Configuration
| Memory size | 4 GB | 6 GB |
|---|---|---|
| Memory type | GDDR5X | GDDR5 |
| Memory bus width | 256 bit | 192 bit |
| Bandwidth | 320.3 GB/s | 192.2 GB/s |
Render Configuration
| Shading units | 1920 | 1280 |
|---|---|---|
| TMUs | 120 | 80 |
| ROPs | 64 | 48 |
| SM count | 15 | 10 |
| Cache L1 | 48 KB (per SM) | 48 KB (per SM) |
| Cache L2 | 2 MB | 1536 KB |
Performance
| Pixel rate | 110.9 GPixel/s | 82.03 GPixel/s |
|---|---|---|
| Texture rate | 208.0 GTexel/s | 136.7 GTexel/s |
| FP16 (half) performance | 104.0 GFLOPS (1:64) | 68.36 GFLOPS (1:64) |
| FP32 (float) performance | 6.655 TFLOPS | 4.375 TFLOPS |
| FP64 (double) performance | 208.0 GFLOPS (1:32) | 136.7 GFLOPS (1:32) |
Dimensions & Outputs
| Slot width | Dual-slot | Dual-slot |
|---|---|---|
| Length | 267 mm 10.5 inches | 250 mm 9.8 inches |
| TDP | unknown | 120 W |
| Suggested PSU | 200 W | 300 W |
| Outputs | No outputs | No outputs |
| Power connectors | 1x 8-pin | 1x 6-pin |
| Board number | PG413 SKU 100 | PG410 SKU 100 |
API Support & Features
| DirectX | 12 (12_1) | 12 (12_1) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.2 | 1.2 |
| CUDA | 6.1 | 6.1 |
| Shader model | 6.4 | 6.4 |
