Ride the 4th Wave: Edge Computing

 By Javier Sanz-Blasco

Let’s flashback to 1999, when I moved from Madrid to London with the help of a paper plane ticket. 5 years later, in 2004, one in every five flights had an e-tickets. By 2008, almost every ticket used by all air carriers worldwide was an e-ticket. 

This is just an example. Technology shifts are happening more quickly now than any time in my 20 year career in the connectivity business. With the advent of Artificial Intelligence and IoT, where employees, customers & devices connect from, how they connect and what they connect make up one of the biggest challenges facing most multi-nationals and e-commerce corporations in 2021.

Wave Number 1. The Mainframe

Not that I was there, but in the 1940s, IBM developed the first general purpose automatic digital computer in conjunction with Harvard University. It has some precursors such as the Colossus, built by the British General Post Office (GPO) to solve a problem around cryptography. Both shared the same principle: vacuum valves.

Wave Number 2. The Personal Computer 

In the late 1970s I was just around, and so were the likes of Microsoft and Apple with their aim to make computers available to everyone. Computers were no longer reserved for geniuses at Harvard and MIT. In the 1990s, I was no genius but I remember cajoling my parents into buying me an unbranded computer for 600_€ where I got initiated with the initial Linux kernels pre 2.0.

Wave Number 3. Cloud. Cloud. Cloud 

Smartphones and cloud were launched at around the same time: Amazon Web Services in 2006 and the iPhone in 2007. Google Cloud and Azure followed a bit later. Fast forward 14 years and there are around 4 billion devices today.

However, most of the processing and queries do not happen locally. Instead, the queries are channelled to Spotify or Netflix or Google or e-commerce servers travel through kilometres of fibre optic wires usually into a warehouse packed full of servers: the data centre. These servers handle our requests, and send us back a trickle of information. Yes, you guessed it: Wave 3 is a return to the mainframe… but this time is as a service. It is the public Cloud, which is not dedicated, like in 1940s with IBM, but shared among different customers. Hosting shared resources has delivered huge economies of scale.

Wave Number 4. Edge Computing

So nowadays a bunch of hyperscalers in a few dozen giant data centres essentially provide modern computing as a service in the same way as utilities companies provide water or electricity. This centralized model works well for technologies of the early 21st century. But it’s too slow for today and tomorrow’s tech breakthroughs: in particular applications that ride the trend towards artificial intelligence or Internet of Things, and that require ultra-real time decisions.

If a driverless vehicle needs to stop suddenly, every second counts. More accurately: every millisecond counts in the decision making process. The same concept may apply to flying drones, or to remote surgery or myriad new 5G-based services.

Suddenly, the Cloud seems to be a little too far away. Computers are once again migrating from a centralized, to a distributed way of working. And edge computing is the key. From the Cloudflare (NYSE: NET) website: Edge computing optimizes Internet devices and web applications by bringing computing closer to the source of the data. This minimizes the need for long distance communications between client and server, which reduces latency and bandwidth usage.

We will require hundreds, not dozens, of smaller datacentres as close as possible to the end users: people or machines. Cloudflare states that 99% of the internet population is within 100ms reach, but is not the only firm moving rapidly into this space. There are big names such as AWS Wavelength, CDN players such as Fastly or startups like Mutable. One or two of them may become the next Google or Microsoft. Who knows? The question is: are you ready?