AWS is now an $8Bn business for Amazon. Could anyone have predicted that in 1997 or even 2007? This week I was talking to a few people about who might be able to disrupt AWS. It won’t be easy. We immediately discounted the old guys like HP, IBM, Dell etc and even Google seems to be a complete also-ran in IaaS. Microsoft Azure might have potential.
PaaS is separate as it’s higher up the stack and makes use of IaaS. I expect Open Source PaaS like OpenShift 3 to grow very quickly in the coming years as it makes deployment and management of Apps incredibly easy. Where PaaS might hurt AWS is in abstracting what IaaS you are using, so as long as the feature set is there, you don’t care where your Apps are running (apart from Geo lag and data protection laws).
One of the people I talked to thinks that AWS will be slayed by a startup who may not even exist yet. He also felt that smart VPSes like DigitalOcean might do it. I’m a huge fan of DO and gained enormous reliability and speed by switching to it from Google Compute Engine/CloudSQL whilst reducing my bill by 2/3rds. But it still feels like an SME/SMB or individual developer solution.
So I’m predicting it’ll be Facebook. Why? Think about the following:
- They understand building things at massive scale on a global basis
- They build their own datacenters
- They built the most energy efficient computing facility ever
- They created the Open Compute project
- They have incubated some incredible Open Source projects like Cassandra and React
- They use PHP. Forget I mentioned that ;-)
- They understand that mobile is eating the world and they own Parse
- They understand real-time messaging at massive scale
- They could do this in their sleep
The big open question is how they do their virtualisation and containerisation. If they have that nailed, then I think they might go for it.
Let’s re-visit this post in 5 years.