Description
Inspur NS5482JD Server Overview
The Inspur NS5482JD is a 4U AI-optimized GPU server built for large-scale model training, deep learning, and AI inferencing. Designed to support up to 16 NVIDIA® GPUs and dual Intel® Xeon® Scalable processors, it delivers exceptional compute density and parallel performance. With a fully interconnected NVLink or PCIe GPU fabric, the NS5482JD accelerates data throughput and AI model training time, making it ideal for advanced AI research institutions and enterprise-level HPC applications.
Key Features
-
4U chassis supporting up to 16x GPUs (A100, H100, or V100)
-
Dual Intel® Xeon® Scalable CPUs (3rd Gen supported)
-
Advanced NVLink or PCIe GPU interconnects
-
Optimized for AI training, inferencing, and HPC modeling
-
Intelligent cooling and redundant power for reliability in dense environments
Inspur NS5482JD Server Specifications
| Component | Specification |
|---|---|
| Form Factor | 4U Rackmount AI Server |
| Processor | Dual Intel® Xeon® Scalable (up to 3rd Gen) |
| Chipset | Intel® C620A or later depending on CPU generation |
| Memory | Up to 2TB DDR4 ECC RDIMM/LRDIMM, 16–32 DIMM slots (CPU dependent) |
| GPU Support | Up to 16x NVIDIA® GPUs (e.g., A100, H100, V100) via PCIe or NVLink |
| GPU Interconnect | PCIe Gen4/Gen5 or NVLink (varies by model/config) |
| Storage Options | Up to 8x 2.5″ NVMe/SATA/SAS hot-swappable drives |
| RAID Support | RAID 0/1/10 supported natively; hardware RAID optional |
| Expansion Slots | Multiple PCIe 4.0/5.0 slots for networking or accelerators |
| Network Interface | Dual 1GbE/10GbE onboard; optional 25/40/100GbE NICs |
| Power Supply | Redundant 3x 3000W Titanium efficiency PSUs |
| Cooling | Optimized for high GPU TDPs with smart airflow and fan zoning |
| Management | IPMI 2.0, Redfish, BMC with KVM support |
| Dimensions | Approx. 448mm (W) × 176mm (H) × ~900mm (D) |
| Certifications | CE, FCC, UL, CCC, RoHS |
Reviews
There are no reviews yet.