Billions of clock speeds run on an edge processor in between network packets, said Chris Sachs, founder of Swim.ai in this episode of The New Stack Makers. The capacity is seemingly timeless. The cloud isn’t meant for processing and analyzing data from nodes in real-time. The data that is processed at the edge from a node requires a fundamental different approach.
Swim.ai’s platform uses software that treats every node as a digital twin. These “actors” are configured as a software kernel, fully configured. They maximize the process clock speeds to use every network packet and CPU cycle to its advantage. It is fully self-contained, a raw compute platform that looks at data from the nodes on the edge.
Swim.ai does real-time data processing by implementing an actor framework that all have their own state -- just enough to maximize all those clock speeds and analyzing it with neural networks that work across intersections, said Swim.ai’s Simon Crosby, who also joined us for the interview.
The capability to process data at the edge allows to create learning environments such as with predicting traffic in a city like Palo Alto, which Sachs and Crosby discuss in particular in this podcast and demonstration.
Watch on YouTube: https://youtu.be/gslqyzZSTt0