Edge Computing Started with the Abacus

Edge computing and edge AI are extremely popular tech buzzwords today.  Just do a quick Google search for this term and you’ll get about 233 million results in less than a few seconds. Yes, there’s an overwhelming amount of information about edge computing and especially the importance of moving AI to the edge.  Taking into consideration the technological leaps and bounds we’ve made in a relatively short period of time, it’s so interesting to reflect on the devices created along the way – each a major stepping stone in the history of computing 

The evolution of computers started with the abacus, a calculating tool (and type of standalone computing machine) used in the near east, Europe, China, and Russia as far back as 2700 BC. Then came the variants of the adding machine originally created by Blaise Pascal and Wilhelm Schickard, from Napier’s bones (1617 AD) to mechanical calculators such as the Calculating Clock (1623 AD). 

 

Calculating Clock created in 1623 AD (Britannica.com)

 

Then came the punched card systems, originally introduced by Jacquard during the turn of the 19th century. The Jacquard loom is important to the history of computers because it is the first machine to use interchangeable punch cards to instruct a machine to perform automated tasks.

Jacquard loom using punched card system (Wikipedia)

I have fond memories of punched cards.  My mom’s side of the family in Doddampalayam, India are weavers and used these Jacquard looms to produce intricately designed sarees. They had no idea about the complexities that went into the design of those punched cards! 

More recently, computers were designed with vacuum tubes, then with transistors and integrated circuits. The IC technology led us to the VLSI (1971 AD) and microprocessor evolution that has defined the massive strides we’ve made in every field. 

Vacuum tubes were invented in 1904 and used through the 1960’s, eventually replaced by transistors (Wikipedia)

As computing capabilities have grown, the network of connecting computing started to gain traction with the ARPANET in the late 1960s to the creation of NSFNET in 1980, an organization that created a backbone of computer networks to support this initiative. 

As the networks grew to popularity, the concept of computing, both from hardware and software aspects, has gone through a multitude of changes. Computing devices were classified by size (supercomputers to micro), functionality (servers to embedded) and data handling (analog, digital, hybrid). 

On the backend, networks also kept evolving, from simple switch circuits to routers to software-defined networks. To support mobility, wireless protocols came into play along with wireless routers and access points.

From the abacus to handheld devices and even nano computing, the history of computing is remarkable.  Yet the technology advances and evolution creates a more complex environment for us. Life was simpler when you only had to access that dumb terminal to connect to your session. Now we’re smartphone-dependent and computing is ubiquitous, meaning there is the computational capability in almost all objects in our environment.  Just look around you. Your refrigerator, home security, and monitoring system, TV, car, watch, fitness band — just to name a few– all have some computational power. This leads us to the Internet of Things and, IoT is about having these objects in our environment all connected to the internet.  

All these IoT devices are becoming the sensory cortex of the world converting analog information into digital data. 

Edge computing is defined as the deployment of data-handling activities or other network operations away from centralized and always-connected network segments (the cloud) and toward individual sources of data capture —  endpoints like laptops, tablets or smartphones. And, the edge”itself is defined by the use case and application it serves. For example, your fitness tracking device with its sensor can be the edge. But when it comes to your fitness stats and visualizing your performance, your mobile device with the fitness tracker app is actually the edge. So “the edge” is actually a continuum vs. a destination. 

However, the computing power is not distributed evenly across the continuum. We have to plan for the lowest available compute and then process accordingly. At Latent AI our mission is to enable ambient computing, regardless of any resource constraints. We’re developing groundbreaking developer tools that greatly reduce the computing resources needed to process AI on the edge, while at the same time are completely hardware-agnostic and seamlessly adapt to the specific needs of the applications. 

Visit us at https://latentai.com to learn more.

Jags Kandasamy, Founder, CEO, LatentAI, Inc.

Photo Credits:  Wikipedia, Brittanica.com


Leave a Reply

Your email address will not be published. Required fields are marked *