TinyML is a technique that shrinks deep learning networks to fit onto small hardware. This process takes edge computing to the extreme by allowing small devices to quickly process data without experiencing latency issues.
Approximately 15.2 million units of TinyML chipsets were shipped two years ago. This number is forecasted to grow to 1.2 billion units this year (a 164x increase.)
TinyML is part of the Edge Computing meta trend.
Edge computing is when data is processed locally (for example, by a nearby device vs. being sent to the cloud.)
Some of the advantages of edge computing include improved privacy, security, latency and load balancing.
Overall, the edge computing market is forecasted to grow 7x by 2028.