Cloud as a Business™: The Economics of CapEx and OpEx

You can start with an analysis of the true benefits of moving critical computing functions into the Cloud. This analysis is largely framed by a straight-up comparison between buying servers, renting portions of servers at co-location (co-lo) facilities, or buying computing and storage capacity in a Public Cloud. The results of this comparison are fairly predictable: buying Cloud capacity is cheaper than renting a part of a server is cheaper than buying your own hardware. A secondary and common element is the ease of scaling an application up or down in the Cloud.

But there are other key elements to any shopping decision for a Cloud deployment, not only related to Cloud versus co-lo versus buy, but also within the diverse range of Cloud providers themselves. Further, in any CapEx/OpEx calculation it’s important to consider capabilities that are truly unique to cloud based deployments that may not show up in traditional CapEx and OpEx calculations. So here’s a quick rundown.

Encourage your customers to think of cloud computing like the difference between a timeshare, a vacation condo shared with some friends, and a standalone vacation home that you own yourself. If you own the vacation home, you have to maintain it and pay for it. If the roof leaks, you pay for it. You bear all costs, even if you only use it twice a year. This is the same as owning your own hardware, running your own data center and computing infrastructure. If you share a vacation condo with a few friends, you have a co-lo facility. It’s a bit better than owning outright yourself, but chances are the condo remains empty some of the time and the maintenance costs are still heavy. With a vacation timeshare, lots of other people help pay for the upkeep of the unit and its almost always occupied.

Like a timeshare, the Cloud allows lots of people to share computing capacity on an as- needed basis, allowing them to increase or dial back computing resources when desired. This type of arrangement can reduce an organization’s capital and operating expenses substantially. Microsoft pegged the number at 80%. A 2011 survey of CIOs in the United Kingdom by IT consultancy Gartner found Cloud savings estimates in IT spending of 50%.

In CapEx, this reduction cuts across the entire data center infrastructure, including servers, storage arrays, software licenses (when needed), routers, and load-balancers.

On the OpEx side, costs shared by Cloud deployments include sys admins, hardware engineers, network engineers, facilities management, electricity, fire protection, and insurance or local and state taxes on facilities. There are other hidden OpEx costs that a Cloud instance can eliminate such as purchasing and acquisition overhead, asset insurance, and business interruption planning and software. Any Cloud provider worth its salt has essentially baked business interruption planning into the mix.

But not all clouds are alike. The recent Amazon AWS outage in the US highlighted some of its architectural weaknesses that adversely impacted customers for days, with aftershocks that will last for months. The architecture of the Cloud is a critical factor in determining stability and reliability. Also, software differentiation between Cloud providers can mean that some Clouds are slower than others. Running on less efficient Clouds costs more because more compute capacity and storage are required. Learn more at our upcoming webinar:

Join KDDI-Joyent for their webinar entitled “The Economics of CapEx and OpEx in the Cloud on Tuesday, June 14.