Have you ever grabbed a hot skillet only to immediately drop it? Do you think your brain had anything to do with this event? If you answered yes, think again. Evolution has worked magic and in order to minimize injury, the spinal cord responded. It’s faster than if the message had to route via the brain and back.
This is the basis behind fog computing. There is a huge efficiency to be gained in putting some intelligence in devices that are closer to the action. Parallels both in natural and human created networks abound. The sheer number of devices that have an IP address – hence a discreet communicable entity- mandate that this type of computer will shortly overtake massive, centralized processes. Here’s why…
A numbers game
In 2015, the number of connected devices is thought to be around 18 billion worldwide. By 2020, that number will escalate to around 50 billion. That is roughly 7 devices connected to every person on the planet. It’s interesting (and easy) to think of a vast, central Borg-like intelligence controlling every device equally sharing information and command structure. In reality (or my view of it) what’s more likely to happen is distributed generic control but mostly smaller groups performing intelligence gathering and response.
That’s how an army functions and how nature functions. Groups of specialized units performing specific functions and responses under general guidance are how things work. An ant colony comes to mind. There are soldier ants, workers, queen and assorted support roles. Each having a role to play, each supporting the collective and each having a sphere of response protocol yet responding to central orders. Think of a bee colony, each bee gathering nectar as programmed but a disturbance in the hive and the collective “sting” order gets issued and it’s all stingers on deck.
Fog Computing Topology
The following chart shows how Fog computing forms a layer between devices and the cloud. More interactive and responsive close to the devices, heavy computing closer to the cloud.
Bigger in what way?
The volume of devices, even if small in computing horsepower will overwhelm the biggest cloud installation by virtue of numbers.
- Bigger in overall computing ability
- Bigger (faster) in terms of response time
- Faster to heal – more modular
- Faster to evolve
Fog computing is distributed computing taken to the nth degree. We would do well to heed the learnings found in nature. Small but powerful. Lots of unintended consequences like expanded security and points of failure etc. If we’re not careful, we’ll be stung.