Inspur EIS800E Half-Width Edge Server Review

It has been some time since we have looked at an Intel Atom C3000 series edge server, but we have one today, the Inspur EIS800E. This is an interesting edge server, but it is also part of a larger family of servers covering a number of scenarios. As we will cover in this review, the EIS800E has a number of expansion options and a form factor that set it apart from its competition. This also provides an opportunity to take a look at the Atom C3000R refresh parts.

Inspur EIS800E Half-Width Edge Server External Overview

The Inspur EIS800E is a half-width server. This is one of the more interesting parts that Inspur has because it is designed around a common platform that covers an array of form factors. The half-width platform is designed both for standalone operation, as we are looking at here, as well as dense rack configuration.

The front of the unit is where most of the base I/O is located. There are four USB 3 ports and a VGA port for local management. For networking, there are both two 10Gbase-T and two SFP+ ports for four 10GbE links in total. There is even an out-of-band management port.

On the side of the EIS800E, we can see that this is more of a standalone configuration. Here we can see various headers for wireless antennas.

We can see a similar layout on the opposite side. These can be used for WiFi, 4G/5G, and more.

At the rear of the unit, we can see a power supply and two expansion slots. Our unit has a full-height PCIe slot as well as an OCP NIC 3.0 slot.

The power supply is an easily swappable design, but since there is only one of these power supplies in the unit there is not the built-in redundancy that two would provide. It is common to see single power supplies in half-width servers and switches. It is less common to see these single PSUs also have an easily serviceable design.

This rear area of the unit is also configurable. For example, there are options to replace this section with 4x 2.5″ or 1x 3.5″ storage.

There is also a super-cool option with four NVIDIA Jetson modules for edge AI inference.

Next, let us get inside the system to see how it is built.