HOME > Product > GPU/Deep Learning > AI high-peMulti GPU for AI&DL >

KG 2208-G2 GPU server

2U rack type, supports 2 full-height and full-length GPU computing cards;

Suitable for deep learning training or inference end applications;

   CPU: Intel® Xeon® Silver 4210 10-core 2.2GHz

   Memory: 16GB DDR4 ECC RDIMM

   GPU: NVIDIA Tesal T4

   Hard Disk: 240GB Data Center SSD

   Network: Intel®i350 dual gigabit network

   Power supply: 1300W 1+1 platinum power supply

   Price : $4800.00
   Basic configuration for reference only

Heterogeneous acceleration
Support 2 Intel® Xeon® Platinum/Gold/Silver/Bronze series scalable processors ;

At the same time, with the help of 2 GPU accelerators CUDA core, Tensor core, RT core and other features of NVIDIA's latest Turing architecture, 

the processing speed is increased by dozens of times compared with the CPU.

Stable and reliable  
Adopt a unique GPU heat dissipation design to ensure the effective heat dissipation of the GPU without frequency reduction;

The power supply module adopts a 1300W platinum power supply and a 1+1 redundancy design, 

which allows the system to not shut down and work without interruption in the event of a power module failure.

Out of the box

The ubuntu operating system, deep learning SDK (CUDA, cuDNN, NCCL), deep learning frameworks including Caffe, TensorFlow, PyTorch, MXNet, 

and all necessary drivers are pre-installed before the factory, and delivered in the form of containers. 

Shorten the software deployment cycle, which takes several days, and use it out of the box to improve work efficiency.

Application scenario
HPC, deep learning, GPU virtualization, graphics rendering, oil and gas exploration and other computationally intensive applications.



2U rack type; 87mm × 447mm × 808mm (H × w × d)


Intel® Xeon®Platinum/Gold/Silver/Bronze series scalable processorTDP 205W


Intel®  C621


10.4 GT/s


24 DIMM sockets, supporting 6TB DDR4 3DS ECC rdimm 2666 / 2933 memory


12 SAS / SATA 12gbps 3.5 inch / 2.5 inch hot swappable hard disks


Two NVIDIA Tesla V100 / P100 / P40 / P4 / T4, RTX 2080ti, RTX Titan


Intel®i350  1000Mb dual port network

PCIe extension

 2 PCI-E 3.0 x16 slots   6 PCI-E 3.0 x8


Aspeed AST2500


IPMI 2.0     

Other interfaces


Power Supply

1300W 1+1 platinum power supply


Windows /10//2012R2/2016/2019

Centos/Redhat 6.9/7.4/7.5; Suse SLES 11sp4/12sp2

Ubuntu 16.04/18.04; Vmware ESXI 6.5/6.7; Xen Server 7.1/7.2

working temperature


Working humidity


SALES:(+86) 135 5249 6467 skype:353043744@qq.com    Email:yj@jpgd.cn

ADD:4th Street, Shangdi, Haidian District, Beijing, China

Inquiry Us Now !

Copyright (c) 2020 jpgpu.com All Rights Reserved