NVIDIA Quadro4 580 XGL vs NVIDIA Tesla P100 DGXS
Processeur graphique
| Nom du GPU | NV18 | GP100 |
|---|---|---|
| Architecture | Celsius | Pascal |
| Fonderie | TSMC | TSMC |
| Taille du processus | 150 nm | 16 nm |
| Transistors | 29 million | 15,300 million |
| Taille des matrices | 65 mm² | 610 mm² |
Carte graphique
| Date de publication | Nov 12th, 2002 | Apr 5th, 2016 |
|---|---|---|
| Génération | Quadro4 XGL | Tesla |
| Production | End-of-life | End-of-life |
| Bus interface | AGP 8x | PCIe 3.0 x16 |
| Critiques | 26 in our database | — |
Vitesse clock
| GPU clock | 300 MHz | — |
|---|---|---|
| Mémoire clock | 200 MHz 400 Mbps effective | 715 MHz 1430 Mbps effective |
| Base clock | — | 1328 MHz |
| Boost clock | — | 1480 MHz |
Mémoire
| Taille de la mémoire | 64 MB | 16 GB |
|---|---|---|
| Type de mémoire | DDR | HBM2 |
| Bus mémoire | 128 bit | 4096 bit |
| Bande passante | 6.400 GB/s | 732.2 GB/s |
Render Config
| Pixel shaders | 2 | — |
|---|---|---|
| Vertex shaders | 0 | — |
| TMUs | 4 | 224 |
| ROPs | 4 | 96 |
| Shading units | — | 3584 |
| Comptage SM | — | 56 |
| Cache L1 | — | 24 KB (per SM) |
| Cache L2 | — | 4 MB |
Performance théorique
| Taux de pixel | 1.200 GPixel/s | 142.1 GPixel/s |
|---|---|---|
| Taux de texture | 1.200 GTexel/s | 331.5 GTexel/s |
| FP16 (half) performance | — | 21.22 TFLOPS (2:1) |
| FP32 (float) performance | — | 10.61 TFLOPS |
| FP64 (double) performance | — | 5.304 TFLOPS (1:2) |
Board Design
| Taille Slot | Single-slot | — |
|---|---|---|
| TDP | unknown | 300 W |
| Suggestions pour le PSU | 200 W | 700 W |
| Sorties | 1x LFH60 | No outputs |
| Connecteurs de puissance | None | None |
Caractéristiques graphiques
| DirectX | 7.0 | 12 (12_1) |
|---|---|---|
| OpenGL | 1.5 | 4.6 |
| OpenCL | — | 3.0 |
| Vulkan | — | 1.2 |
| Pixel shader | 1.1 | — |
| Vertex shader | 1.1 | — |
| CUDA | — | 6.0 |
| Shader model | — | 6.4 |
