From a deck that I presented at the SIIA “All About Mobile” conference in November 2011 in San Francisco. It starts with the usual set of slides on the recent history of mobility (and I will keep presenting them until I see no more “I had no clue” faces in an audience), and then goes deeper into Moore’s Law and how we see it continuing for cell phones.
An additional journey back in history to the early days of the industrialization and electricity. Companies had to generate their own power (by using wind, water, animals, etc.) and Burden’s Wheel is a good example of one big, giant monolithic effort to do so. Along came Tesla and Westinghouse, and the first power plant “Adam’s Plant” was able to provide about 3x the power, but over a much further distance, and to multiple customers. The concept of an electric utility was born, and what we saw happening was the fall of “enterprise power generation”.
Fast forward to 1969, and Douglas Parkhill and John McCarthy came up with the concept of the “Computer Utility”. Today we see multi-$B investments into public cloud infrastructures. In very simple terms, if history in the utility industry is any indicator, we will see enterprise clouds disappear. And as cloud infrastructures scale and get more efficient, and the price of computing goes down (Moore’s Law), developers will find a way to use and instrument that computing power, and make it consumable to enterprises and consumers, which gets us to Jevons’ Paradox.
Jevons observed how consumption of energy in England went up as coal power plants got more efficient. All the way to today where we keep the lights on in our homes 24/7, and darkness has actually become a scarce good in some metropolitan areas. Switching to enterprise computing and looking at BEA data on IT assets for the past four decades, we see that prices for IT assets are falling, whereas other assets follow an inflationary path. And as computing gets cheaper, enterprises consume more and more of it (and you can argue so do consumers, aka “Consumerization of IT”).
What is striking that with the arrival of the public Internet in ‘90-95 and web companies like Yahoo and Amazon, the mix in consumption is shifting: it’s increasingly going towards software, up from a SW:HW ratio of roughly 1:1 over three decades, to now 3:1. So today, for every $1 spent on hardware, enterprises spend $3 on software. Hence, it seems like enterprises are making use of the public cloud, which would explain the rise of SaaS companies, such Salesforce, SuccessFactors and also Amazon’s AWS.
And as the rise of smartphones is only beginning, enterprise mobility will likely drive the trend of an increasing SW:HW ratio further up. I wouldn’t be surprised to see the mix go to 10:1 in the next five years as smartphones proliferate and the amount of on-deck and off-deck computing power available to a single device is growing exponentially, the concept of “accelerated acceleratio