NVIDIA A16 PCIe vs NVIDIA A2
Processeur graphique
| Nom du GPU | GA107 | GA107 |
|---|---|---|
| Architecture | Ampere | Ampere |
| Fonderie | Samsung | Samsung |
| Taille du processus | 8 nm | 8 nm |
| Transistors | unknown | unknown |
| Taille des matrices | unknown | unknown |
Carte graphique
| Date de publication | Apr 12th, 2021 | Nov 10th, 2021 |
|---|---|---|
| gpu.details.availability | 2021 | — |
| Génération | Tesla | Quadro |
| Production | Active | Active |
| Bus interface | PCIe 4.0 x16 | PCIe 4.0 x8 |
Vitesse clock
| Base clock | 885 MHz | 1440 MHz |
|---|---|---|
| Boost clock | 1695 MHz | 1770 MHz |
| Mémoire clock | 1812 MHz 14.5 Gbps effective | 1563 MHz 12.5 Gbps effective |
Mémoire
| Taille de la mémoire | 16 GB | 16 GB |
|---|---|---|
| Type de mémoire | GDDR6 | GDDR6 |
| Bus mémoire | 128 bit | 128 bit |
| Bande passante | 231.9 GB/s | 200.1 GB/s |
Render Config
| Shading units | 2560 | 1280 |
|---|---|---|
| TMUs | 80 | 40 |
| ROPs | 48 | 32 |
| Comptage SM | 20 | 10 |
| Tensor cores | 80 | 40 |
| RT cores | 20 | 10 |
| Cache L1 | 128 KB (per SM) | 128 KB (per SM) |
| Cache L2 | 2 MB | 2 MB |
Performance théorique
| Taux de pixel | 81.36 GPixel/s | 56.64 GPixel/s |
|---|---|---|
| Taux de texture | 135.6 GTexel/s | 70.80 GTexel/s |
| FP16 (half) performance | 8.678 TFLOPS (1:1) | 4.531 TFLOPS (1:1) |
| FP32 (float) performance | 8.678 TFLOPS | 4.531 TFLOPS |
| FP64 (double) performance | 271.2 GFLOPS (1:32) | 70.80 GFLOPS (1:64) |
Board Design
| Taille Slot | Dual-slot | Single-slot |
|---|---|---|
| Longueur | 267 mm 10.5 inches | — |
| Taille | 112 mm 4.4 inches | — |
| TDP | 250 W | 60 W |
| Suggestions pour le PSU | 600 W | 250 W |
| Sorties | No outputs | No outputs |
| Connecteurs de puissance | 8-pin EPS | None |
| Board number | — | PG179 SKU 220 |
Caractéristiques graphiques
| DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.2 | 1.2 |
| CUDA | 8.6 | 8.6 |
| Shader model | 6.6 | 6.6 |
