NVIDIA GRID A100A vs NVIDIA Quadro K4000
Processeur graphique
| Nom du GPU | GA100 | GK106 |
|---|---|---|
| Architecture | Ampere | Kepler |
| Fonderie | TSMC | TSMC |
| Taille du processus | 7 nm | 28 nm |
| Transistors | 54,200 million | 2,540 million |
| Taille des matrices | 826 mm² | 221 mm² |
Carte graphique
| Date de publication | May 14th, 2020 | Mar 1st, 2013 |
|---|---|---|
| Génération | GRID | Quadro |
| Production | Active | End-of-life |
| Bus interface | PCIe 4.0 x16 | PCIe 2.0 x16 |
| Prix de lancement | — | 1,269 USD |
| Critiques | — | 26 in our database |
Vitesse clock
| Base clock | 900 MHz | — |
|---|---|---|
| Boost clock | 1005 MHz | — |
| Mémoire clock | 1215 MHz 2.4 Gbps effective | 1404 MHz 5.6 Gbps effective |
| GPU clock | — | 810 MHz |
Mémoire
| Taille de la mémoire | 48 GB | 3 GB |
|---|---|---|
| Type de mémoire | HBM2E | GDDR5 |
| Bus mémoire | 6144 bit | 192 bit |
| Bande passante | 1,866 GB/s | 134.8 GB/s |
Render Config
| Shading units | 6912 | 768 |
|---|---|---|
| TMUs | 432 | 64 |
| ROPs | 192 | 24 |
| Comptage SM | 108 | — |
| Tensor cores | 432 | — |
| Cache L1 | 192 KB (per SM) | 16 KB (per SMX) |
| Cache L2 | 48 MB | 384 KB |
| Comptage SMX | — | 4 |
Performance théorique
| Taux de pixel | 193.0 GPixel/s | 12.96 GPixel/s |
|---|---|---|
| Taux de texture | 434.2 GTexel/s | 51.84 GTexel/s |
| FP16 (half) performance | 55.57 TFLOPS (4:1) | — |
| FP32 (float) performance | 13.89 TFLOPS | 1,244 GFLOPS |
| FP64 (double) performance | 6.947 TFLOPS (1:2) | 51.84 GFLOPS (1:24) |
Board Design
| Taille Slot | IGP | Single-slot |
|---|---|---|
| TDP | 400 W | 80 W |
| Suggestions pour le PSU | 800 W | 250 W |
| Sorties | No outputs | 1x DVI2x DisplayPort |
| Connecteurs de puissance | None | 1x 6-pin |
| Longueur | — | 241 mm 9.5 inches |
| Taille | — | 111 mm 4.4 inches |
| Board number | — | P2030 |
Caractéristiques graphiques
| DirectX | — | 12 (11_0) |
|---|---|---|
| OpenGL | — | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | — | 1.1 |
| CUDA | 8.0 | 3.0 |
| Shader model | — | 5.1 |
