Last week I decided to put the statement to a test with this formula:
100 - [size of public cloud market] / ([size of servers, storage, security & virtualization market] * [average hardware refresh cycle] - [size of public cloud market]) * 100
Different analysts provide different numbers, so you can get slightly different results depending on the data sources you use. But like we used to say at Dell, the results should be “directionally correct”. So here it goes using primarily Gartner as a data source:
100 - $10B / ($140B * 5 - $10B) * 100
The result: 98.5%. Or stated in the converse, only 1.5% of infrastructure has moved into the public cloud. So Google wasn’t far off after all.
Using estimates of AWS’s Compound Annual Growth Rate (estimated between 45-55%), we can assume the percentage of infrastructure moving into the cloud has been quietly accelerating since 2007. I can also report from the frontlines of the public cloud the following change in attitudes in the enterprise over the last year:
- Recognition the increasing performance of IaaS services (e.g. Amazon’s new instance type families) now allows almost any workload to be run in the cloud.
- Realization as to the scope of their internal public cloud investment driven by line of business adoption.
- Increasing acceptance of the security / reliability of the public cloud.
- Increasing focus on shutting down entire data centers contain 100s/1000s of applications.
- Decreasing interest in continuing investment in private clouds managed by internal IT.
So we stand in the early phases of a seismic disruption in our industry that will likely take 20+ years to complete. It’s like the internet in 1994: great promise, lots of buzz, but just getting started. For all that has been accomplished, many substantial challenges still need to be solved to achieve the tipping point: the day when more infrastructure is run in the public cloud than is running outside of it.
But that day is coming. Just 48.5%+ to go.