Thursday 21 January 2021

What’s the Difference Between PoE and PoE+?

Original Post - Press Office, VersaLogic Corporation, 12/03/20


The team at VersaLogic are constantly pushing boundaries and exploring new technology. Its one of the many reasons we enjoy working with them.

In December they published a detailed blog post providing an overview of PoE (Power over Ethernet), and the difference between PoE and PoE+. It explains power sourcing and the various ways that devices in the system can be powered. The original blog post goes into much more detail, you can read it here.

PoE, short for Power over Ethernet, is a way of powering peripheral devices, such as cameras or wireless access points, without the need for a separate power cable or local power source. All of the power needed to operate the device is provided over the standard Ethernet cable.

PoE comes in a variety of flavors and the related standards have been developed within the overall remit of the IEEE 802.3 Ethernet Working Group. The original definition of PoE envisaged the Power Sourcing Equipment (PSE) would deliver 15.4W and, after line drops, 12.95 W would be available to the Powered Device (PD). PoE+ takes this up to 30W from the PSE with 25.5W available at the PD. Higher levels are also possible with PoE++ which specifies up to 100W at the PSE.

The terms PoE and PoE+ are widely used, but the terms Type 1 (PoE) and Type 2 (PoE+) are also used. In addition, there are also 5 defined power classes. During power-up, the PSE determines the power classification of the PD. To avoid damage to non-PoE enabled devices, the PSE will only supply power to PoE enabled devices.

There are many types of PDs such as cameras, wireless access points, VOIP phones, light fixtures, and sensors. Their need for PoE versus POE+ is based strictly on their total power requirements (either above or below 12.95W).

Whether a wireless access point requires a PoE or PoE+ may depend on the environmental conditions. When operating in environments of > -20°C, PoE power is sufficient. However, when the temperature plummets into the range -50° to -20°C an internal heater is required, driving the need for PoE+ power.

Industrial / embedded computers may have PoE ports built-in, or they can often use some type of PoE expansion board. VersaLogic’s “Grizzly” computer board is an example of a high-performance system that includes built-in PoE ports.

For a deeper dive into PoE, Microsemi has a white paper that provides more detail. If you'd like to know more about VersaLogic’s range of products with PoE options please get in touch with one of our knowledgeable engineers.

 Read the Full Blog Post

Friday 15 January 2021

WEBINAR: Leveraging the SOSA™ Architecture to Enable AI Solutions for EW/ISR Applications

In a previous blog post we discussed how the new SOSA architecture was going to be big news for C4ISR systems. Fast forward a year, and as AI has become more prevalent in emerging computing solutions, it is no surprise that it has become an integral part of the conversation when discussing the implementations within the world of SOSA (Sensor Open System Architecture ).

We are very excited to be able to share this webinar with you, which will examine how AI is being designed into SOSA modules, with a specific focus on RF and AI solutions.



DATE: Wednesday 27th January, 2021
TIME: 7pm GMT / 1pm CST
DURATION: 60 minutes

Register here: Webinar Registration 

Speakers:
Chris Fadeley - Software Engineering Manager, EIZO Rugged Solutions
Nigel Forrester - Director of Business Development, Concurrent Technologies
Ken Grob - Director of Embedded Computing, Elma Electronic

Moderator: John Keller - Editor-in-Chief, Military & Aerospace Electronics

Alongside speakers from Concurrent Technologies and Elma Electronic, Chris Fadeley, Software Engineering Manager at EIZO Rugged Solutions will discuss in-depth how different components of a modular open architecture system can deliver convincing artificial intelligence solutions. His discussion will cover a number of areas, including:

  • Quick briefing on Traditional Algorithms vs Deep Learning
  • High level explanation covering how Deep Learning is performed in a real world environment (Training and Inference)
  • Discussion about the current rollout in the Embedded market and the SOSA-aligned options available for Deep Learning hardware.
  • Overview of the different software stacks available and NVIDIA specific support.

The webinar will cover key areas, including where GPGPUs excel along with their software options; what advantages CPU and FPGA designs have in AI, and what challenges can occur when integrating a system aligned to SOSA to meet the inevitable cooling challenges. 

Registration is free, we are sure the content will be incredibly useful for those who are looking at the implementation of SOSA into RF and AI solutions. If you aren't able to make it on the day please get in touch with us, as we will be able to supply a link to the recording after the event.