The growing list of applications for artificial intelligence (AI) covers a wide area including such things as face recognition, image processing, natural language processing, military decision making, robotic control, natural language translation, data mining - the list goes on. When AI service is needed at the “edge,” there is inevitable latency that can compromise in-time response to real situations.
As an example, in 2018, only 10% of generated data was created and processed at the edge. By 2025, it is expected to grow to 75%. Processing simply must go out to the edge to meet it without having to transport it to another server environment.