Deutsch
English
Main features
| Produktbeschreibung | NVIDIA A30 - GPU-Rechenprozessor - A30 - 24 GB |
| Type | Graphics card |
| Gerätetyp | GPU-Rechenprozessor |
| Bus type | PCI Express 4.0 x16 |
| Grafik-Engine | NVIDIA A30 |
| Speicher | 24 GB HBM2 |
| CUDA-Kerne | 3804 |
| Speicherschnittstelle | 3072-Bit |
| Ohne Lüfter | Ja |
| Abmessungen (Breite x Tiefe x Höhe) | 26.77 cm x 11.115 cm |
General
| Gerätetyp | GPU-Rechenprozessor - ohne Lüfter |
| Bus type | PCI-Express |
| Grafik-Engine | NVIDIA A30 |
| CUDA-Kerne | 3804 |
| Besonderheiten | 5,2 TFlops Spitzenwert für Gleitkomma-Performance mit doppelter Präzision, 10,3 Tflops Spitze Floting Point Leistung mit einzelner Präzision, Dual-Slot-Breite, NVIDIA Tensor Core, NVIDIA Ampere GPU-Technologie, MIG-Technologie (Multi-Instance GPU) |
Memory
| Grösse | 24 GB |
| Technologie | HBM2 |
| Bus width | 3072-Bit |
| Bandbreite | 933 GBps |
Systemanforderungen
| Zusätzliche Anforderungen | 8-poliger CPU-Stromanschluss |
Verschiedenes
| Leistungsaufnahme im Betrieb | 165 Watt |
| Tiefe | 26.77 cm |
| Höhe | 11.115 cm |
| Weight | 1.024 kg |
Herstellergarantie
| Service und Support | Begrenzte Garantie - 3 Jahre |
Classification
| eClass-10.1-ID | 19050490 |
| eClass-11.1-ID | 19050490 |
| eClass-6.2-ID | 19050490 |
| eClass-7.1-ID | 19050490 |
| eClass-8.1-ID | 19050490 |
| eClass-9.1-ID | 19050490 |
| eClass-ID | 19030309 |
| UNSPSC 20.0601 Code | 43201401 |
NVIDIA A30 Tensor Core GPU is the most versatile mainstream compute GPU for AI inference and mainstream enterprise workloads. Powered by NVIDIA Ampere architecture Tensor Core technology, it supports a broad range of math precisions, providing a single accelerator to speed up every workload. Built for AI inference at scale, the same compute resource can rapidly re-train AI models with TF32, as well as accelerate high-performance computing (HPC) applications using FP64 Tensor Cores. Multi-Instance GPU (MIG) and FP64 Tensor Cores combine with fast 933 gigabytes per second (GB/s) of memory bandwidth in a low 165W power envelope, all running on a PCIe card optimal for mainstream servers.
