Artificial AI Intelligent 4U Rack-mounted Multi-graphics GPU Server Dual-channel Cloud Storage Cloud Service Host Platform

PARAMETERS：

1.  Intel 1/2/3/4 Xeon Scalable Series CPUs

2.  Support VGA display output

3.  Supports 24 * DIMM DDR4 RECC memory

4.  2 * Intel Gigabit Ethernet card (optional 10 Gigabit Ethernet card)

5.  8 hot plug and hot unplug 2.5-inch SATA/SAS hard drive slots

6.  Support IPMI remote management

7.  The whole machine supports a maximum of 8/10 GPU cards

8.  Size: 798 * 435 * 176mm

HOST CHARACTERISTICS：

Supports Intel Xeon Scalable Series processors (1ce Lake), supports 2 high-power CPU motherboards, native 2 M.2 interfaces for ultra-high computing power, and has been comprehensively improved in terms of kernel frequency cache interconnection channels. It adopts DDR4 technology to support 2TB of memory expansion capability, which can be expanded on demand to meet the application requirements of ultra-high computing power such as digital currency, cloud computing, big data, virtual computing, etc. The 4U system can support up to 8 dual full height GPU cards, compatible with NVIDIA and AMD graphics cards, providing multiple power supply methods without changing accessories, providing unparalleled excellent architecture for pursuing maximum computer performance

APPLICATION SCENARIOS：

Can support NVIDIA RTX4090 A100 A800 H100 H800 and other GPU graphics cards, suitable for applications such as artificial intelligence deep learning training, reasoning, mathematical modeling, high-performance computing, etc

TECHNICAL SPECIFICATIONS：