The edge is modular
There are exciting times ahead for the edge. If analysts’ predictions are true, we will see a massive growth at a CAGR of 54.0%[i] in the edge computing market over the next few years. Micro data centers will be located close to the infrastructure to address real-time issues in nearly all embedded markets. The cry for massive edge computing performance is coming from OEMs in markets as varied as industrial, energy & utilities, healthcare, transportation & logistics as well as retail and smart buildings & cities. The embedded computing industry already serves all these markets with low-power, ruggedized platforms that offer long term availability for harsh environments. But until today, there was a gap in processing performance for the server class, which may also include AI. The next generation of rugged mid-range server processors will address this gap, and embedded computing vendors will integrate these new processors on embedded platforms, offering nearly the same, well-known benefits for OEMs as in the past. There will be only one difference: These platforms will have brand new server-grade functionalities, including all required RAMS (Reliability, Availability, Maintainability, and Safety) features such as IPMI interfaces and functionalities for out-of-band management. It is obvious that existing Computer-on-Module standards can neither cater for these server-grade functionalities nor support the latest high-speed interfaces such as 25 GbE or PCIe Gen 4 and Gen 5.
