I came across this “short” (25 minutes or so) video by Andreessen Horowitz partner Peter Levine.
I don’t like people that announce “death of” or “end of” or similar (see Mainframes, COBOL, even tape for a simple explanation why).
However flawed the thinking, there are some very good points he is making (and a very good explanation) of edge computing.
We won’t get rid of cloud computing (a distributed computing model) to adopt edge computing – but we will evolve and improve cloud computing as implemented today to embrace end nodes as part of the cloud. He says this about halfway into the video but then ends up enamored of the idea of ending cloud computing (which would bring up the question on how these “edge nodes” will actually have their marching orders and coordination if not via central infrastructure – at least today).
The topic sets up one of my favorite topics: massively parallel systems (I still today, after many books on the topic, favor Mitch Resnick’s book: “Turtles, Termites, and Traffic Jams” as an easy introduction).
Wanted to share the video as a starting point in the infrastructure topic that is so underhyped (yay) but has massive (half-pun intended) opportunities as we embrace cloud computing (real cloud, not hosted or legacy SaaS solutions as we have deployed in the past).