I recently went to Amazon’s annual AWS (Amazon Web Services) conference. The conclusions:
- Cloud infrastructure is set to take off and much bigger deal than many appreciate. It is not just “renting” compute and storage. It enables a technically innovative, “better” model. Much more efficient. Effectively breaking the link between software and the hardware running it (various caveats here). Its like the early days of electricity shifting from local generators at individual factories to a liquid “pool” of grid power. The resulting productivity gains were major.
- Amazon’s AWS looks set to dominate this market. It was just plain obvious that AWs is past the tipping point and well ahead of peers. And it is a huge market – hundreds of billions of $$. That means they could be incubating another dominant tech titan on par with a Microsoft, IBM, Google, etc….
- Even if Amazon falters, this looks like a natural 2-3 player market. Experience curve effects and economies of scale make it so. The electric grid metaphor is apt here too. So get ready for a world where much of the world’s compute infrastructure is owned and operated by a few huge players.
- Cloud is a sinkhole for much of the traditional IT hardware and software industries. Incremental new computing load is shifting to the cloud. The money will follow that load. Cloud providers don’t buy from traditional IT vendors. So traditional IT revenues will drain away – slowly (at first) but inexorably.
- Cloud computing is a major productivity gain for those who rely on IT – ie. pretty much everyone. It will particularly advantage new/nimble players. The more/faster you build/shift your IT onto a cloud platform, the greater the advantage. Older companies based on decades of “self-generated power” software code layers upon layers will be disadvantaged. Subtle effects, but real.
Scary stuff.
Lets start by following the money. Amazon’s AWS business has gone from basically zero to aboout $5 billion in annual service revenue in just 4 years (since 2010). That is the same revenues as a Salesforce.com. In 4 years.
The underlying volume growth has been roughly doubling, driving year over year revenue growth of @45%, 58%, and 64% going back.
But that $5 billion is still just a tiny pimple on the a** of the world’s IT spend (which is estimated at roughly $1 trillion in what everyone knows is a wild-guess statistic). And cloud computing addresses a HUGE part of that $1T market.
That first $5 billion has also siphoned away from traditional IT spending worryingly fast. Starting with software development loads, but now carrying real-world, mission critical load. Like much of Netflix’s IT infrastructure. Yup – that Netflix. And ALL of Coca-Cola’s @1,000 consumer-facing websites. Yup – that Coca-Cola.
AWS is still in the pimple stage, but the underlying power and breadth are volcanic. As the flows grow, there’s likely a pretty huge landmass here by the time its all over. A few years from now, AWS could be drawing away an incremental $5 billion a year (vs the @$1.5 billion they’ll pull in 2014). These sorts of trends tend to accelerate. With such a huge reservoir behind it, small leaks in the dam will just get bigger. Eventually the barriers just collapses.
And basically every penny of that spend will be coming out of the pockets of someone else. The “traditional” IT hardware and software vendors are facing a lot of pain. And, eventually, extinction. We live in interesting times.
More to follow.
