NVIDIA Tesla P4 vs NVIDIA Tesla P40
Graphics Processor
| GPU model | GP104 | GP102 |
|---|---|---|
| GPU variant | GP104-895-A1 | — |
| Architecture | Pascal | Pascal |
| Foundry | TSMC | TSMC |
| Manufacturing process | 16 nm | 16 nm |
| Number of transistors | 7,200 million | 11,800 million |
| Die size | 314 mm² | 471 mm² |
Graphics Card
| Release date | Sep 13th, 2016 | Sep 13th, 2016 |
|---|---|---|
| Generation | Tesla | Tesla |
| Production | End-of-life | End-of-life |
| Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
| Launch price | — | 5,699 USD |
Clocks
| Base clock | 886 MHz | 1303 MHz |
|---|---|---|
| Boost clock | 1114 MHz | 1531 MHz |
| Memory clock | 1502 MHz 6 Gbps effective | 1808 MHz 14.5 Gbps effective |
Memory Configuration
| Memory size | 8 GB | 24 GB |
|---|---|---|
| Memory type | GDDR5 | GDDR5X |
| Memory bus width | 256 bit | 384 bit |
| Bandwidth | 192.3 GB/s | 694.3 GB/s |
Render Configuration
| Shading units | 2560 | 3840 |
|---|---|---|
| TMUs | 160 | 240 |
| ROPs | 64 | 96 |
| SM count | 20 | 30 |
| Cache L1 | 48 KB (per SM) | 48 KB (per SM) |
| Cache L2 | 2 MB | 3 MB |
Performance
| Pixel rate | 71.30 GPixel/s | 147.0 GPixel/s |
|---|---|---|
| Texture rate | 178.2 GTexel/s | 367.4 GTexel/s |
| FP16 (half) performance | 89.12 GFLOPS (1:64) | 183.7 GFLOPS (1:64) |
| FP32 (float) performance | 5.704 TFLOPS | 11.76 TFLOPS |
| FP64 (double) performance | 178.2 GFLOPS (1:32) | 367.4 GFLOPS (1:32) |
Dimensions & Outputs
| Slot width | Single-slot | Dual-slot |
|---|---|---|
| Length | 168 mm 6.6 inches | 267 mm 10.5 inches |
| TDP | 75 W | 250 W |
| Suggested PSU | 250 W | 600 W |
| Outputs | No outputs | No outputs |
| Power connectors | None | 1x 6-pin + 1x 8-pin |
| Width | — | 111 mm 4.4 inches |
| Board number | — | PG610 SKU 200 |
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 |
