NVIDIA Tesla D870 vs NVIDIA Tesla K40m
Processeur graphique
| Nom du GPU | G80 | GK110B |
|---|---|---|
| Architecture | Tesla | Kepler |
| Fonderie | TSMC | TSMC |
| Taille du processus | 90 nm | 28 nm |
| Transistors | 681 million | 7,080 million |
| Taille des matrices | 484 mm² | 561 mm² |
Carte graphique
| Date de publication | May 2nd, 2007 | Nov 22nd, 2013 |
|---|---|---|
| Génération | Tesla | Tesla |
| Production | End-of-life | End-of-life |
| Prix de lancement | 7,499 USD | 7,699 USD |
| Bus interface | PCIe 1.0 x16 | PCIe 3.0 x16 |
Vitesse clock
| GPU clock | 600 MHz | — |
|---|---|---|
| Shader clock | 1350 MHz | — |
| Mémoire clock | 800 MHz 1600 Mbps effective | 1502 MHz 6 Gbps effective |
| Base clock | — | 745 MHz |
| Boost clock | — | 876 MHz |
Mémoire
| Taille de la mémoire | 1536 MB | 12 GB |
|---|---|---|
| Type de mémoire | GDDR3 | GDDR5 |
| Bus mémoire | 384 bit | 384 bit |
| Bande passante | 76.80 GB/s | 288.4 GB/s |
Render Config
| Shading units | 128 | 2880 |
|---|---|---|
| TMUs | 32 | 240 |
| ROPs | 24 | 48 |
| Comptage SM | 16 | — |
| Cache L2 | 96 KB | 1536 KB |
| Comptage SMX | — | 15 |
| Cache L1 | — | 16 KB (per SMX) |
Performance théorique
| Taux de pixel | 14.40 GPixel/s | 52.56 GPixel/s |
|---|---|---|
| Taux de texture | 38.40 GTexel/s | 210.2 GTexel/s |
| FP32 (float) performance | 345.6 GFLOPS | 5.046 TFLOPS |
| FP64 (double) performance | — | 1.682 TFLOPS (1:3) |
Board Design
| Taille Slot | Dual-slot | Dual-slot |
|---|---|---|
| TDP | 520 W | 245 W |
| Suggestions pour le PSU | 900 W | 550 W |
| Sorties | No outputs | No outputs |
| Longueur | — | 267 mm 10.5 inches |
Caractéristiques graphiques
| DirectX | 11.1 (10_0) | 12 (11_1) |
|---|---|---|
| OpenGL | 3.3 | 4.6 |
| OpenCL | 1.1 (1.0) | 3.0 |
| Vulkan | — | 1.1 |
| CUDA | 1.0 | 3.5 |
| Shader model | 4.0 | 5.1 |
