By IT Brew Staff
less than 3 min read
Definition:
The term fog computing—first used by Cisco’s researchers in 2012—defines a platform providing compute, storage, and networking services between a physical device and a data center. The fog “nodes” that collect and process data—switches, routers, and connected servers—can be applied to any environment with a network connection.
The concept of fog computing emerged in the 2010s with the rise of the internet of things (IoT) and constantly monitoring “smart” devices, with tech professionals wrestling with the practicality of sending all that data through a cloud environment. Many IoT-led processes, like providing alerts for important manufacturing processes, require quick action and low latency; time spent traveling to a cloud platform could potentially slow down automated decisions and actions.
With fog computing, the intermediate, closer-to-the-ground node devices handle the processes at higher speeds. Instead of sending data to a cloud environment, a nearby machine—like a vehicle computer or roadside unit—processes the data supporting “smart” IoT-connected decisions.
Aside from fast responses, a fog model potentially increases security and privacy (with data staying close to home). Disadvantages of fog computing include the maintenance required for any new hardware, like industrial PCs and servers.
The concept of local processing and storage remains relevant today, as many IT pros now debate the pros and cons of having nearby compute for AI. AI fog computing runs artificial intelligence on decision-making machines that sit between an on-the-ground device (like a security camera) and a data center in the cloud.
And if you want to get even closer to the ground than fog, there’s the edge. Edge computing refers to data processing, storage, and networking capabilities at the data’s location. Edge AI runs inference on local hardware, not making requests to the cloud or data center.
