NVIDIA Tesla M40 vs NVIDIA Tesla P4
Processeur graphique
| Nom du GPU | GM200 | GP104 |
|---|---|---|
| GPU variant | GM200-895-A1 | GP104-895-A1 |
| Architecture | Maxwell 2.0 | Pascal |
| Fonderie | TSMC | TSMC |
| Taille du processus | 28 nm | 16 nm |
| Transistors | 8,000 million | 7,200 million |
| Taille des matrices | 601 mm² | 314 mm² |
Carte graphique
| Date de publication | Nov 10th, 2015 | Sep 13th, 2016 |
|---|---|---|
| Génération | Tesla | Tesla |
| Production | End-of-life | End-of-life |
| Bus interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Vitesse clock
| Base clock | 948 MHz | 886 MHz |
|---|---|---|
| Boost clock | 1112 MHz | 1114 MHz |
| Mémoire clock | 1502 MHz 6 Gbps effective | 1502 MHz 6 Gbps effective |
Mémoire
| Taille de la mémoire | 12 GB | 8 GB |
|---|---|---|
| Type de mémoire | GDDR5 | GDDR5 |
| Bus mémoire | 384 bit | 256 bit |
| Bande passante | 288.4 GB/s | 192.3 GB/s |
Render Config
| Shading units | 3072 | 2560 |
|---|---|---|
| TMUs | 192 | 160 |
| ROPs | 96 | 64 |
| Comptage SMM | 24 | — |
| Cache L1 | 48 KB (per SMM) | 48 KB (per SM) |
| Cache L2 | 3 MB | 2 MB |
| Comptage SM | — | 20 |
Performance théorique
| Taux de pixel | 106.8 GPixel/s | 71.30 GPixel/s |
|---|---|---|
| Taux de texture | 213.5 GTexel/s | 178.2 GTexel/s |
| FP32 (float) performance | 6.832 TFLOPS | 5.704 TFLOPS |
| FP64 (double) performance | 213.5 GFLOPS (1:32) | 178.2 GFLOPS (1:32) |
| FP16 (half) performance | — | 89.12 GFLOPS (1:64) |
Board Design
| Taille Slot | Dual-slot | Single-slot |
|---|---|---|
| Longueur | 267 mm 10.5 inches | 168 mm 6.6 inches |
| TDP | 250 W | 75 W |
| Suggestions pour le PSU | 600 W | 250 W |
| Sorties | No outputs | No outputs |
| Connecteurs de puissance | 1x 6-pin + 1x 8-pin | None |
| Board number | PG600 SKU 202 | — |
Caractéristiques graphiques
| DirectX | 12 (12_1) | 12 (12_1) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.1 | 1.2 |
| CUDA | 5.2 | 6.1 |
| Shader model | 6.4 | 6.4 |
