NVIDIA P104-100 vs NVIDIA Tesla M4
Graphics Processor
| GPU model | GP104 | GM206 |
|---|---|---|
| GPU variant | GP104-100-A1 | — |
| Architecture | Pascal | Maxwell 2.0 |
| Foundry | TSMC | TSMC |
| Manufacturing process | 16 nm | 28 nm |
| Number of transistors | 7,200 million | 2,940 million |
| Die size | 314 mm² | 228 mm² |
Graphics Card
| Release date | Dec 12th, 2017 | Nov 10th, 2015 |
|---|---|---|
| Generation | Mining GPUs | Tesla |
| Production | End-of-life | End-of-life |
| Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Clocks
| Base clock | 1607 MHz | 872 MHz |
|---|---|---|
| Boost clock | 1733 MHz | 1072 MHz |
| Memory clock | 1251 MHz 10 Gbps effective | 1375 MHz 5.5 Gbps effective |
Memory Configuration
| Memory size | 4 GB | 4 GB |
|---|---|---|
| Memory type | GDDR5X | GDDR5 |
| Memory bus width | 256 bit | 128 bit |
| Bandwidth | 320.3 GB/s | 88.00 GB/s |
Render Configuration
| Shading units | 1920 | 1024 |
|---|---|---|
| TMUs | 120 | 64 |
| ROPs | 64 | 32 |
| SM count | 15 | — |
| Cache L1 | 48 KB (per SM) | 48 KB (per SMM) |
| Cache L2 | 2 MB | 1024 KB |
| SMM count | — | 8 |
Performance
| Pixel rate | 110.9 GPixel/s | 34.30 GPixel/s |
|---|---|---|
| Texture rate | 208.0 GTexel/s | 68.61 GTexel/s |
| FP16 (half) performance | 104.0 GFLOPS (1:64) | — |
| FP32 (float) performance | 6.655 TFLOPS | 2.195 TFLOPS |
| FP64 (double) performance | 208.0 GFLOPS (1:32) | 68.61 GFLOPS (1:32) |
Dimensions & Outputs
| Slot width | Dual-slot | Single-slot |
|---|---|---|
| Length | 267 mm 10.5 inches | — |
| TDP | unknown | 50 W |
| Suggested PSU | 200 W | 250 W |
| Outputs | No outputs | No outputs |
| Power connectors | 1x 8-pin | — |
| Board number | PG413 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 | 1.1 |
| CUDA | 6.1 | 5.2 |
| Shader model | 6.4 | 6.4 |
