Get your FREE 30-day trial.
Please complete all fields.
Trying to figure out the total cost of ownership (TCO) for your compute infrastructure is no easy task. With the proliferation of public, private, and hybrid clouds, not to mention the ever growing number of Buzzword-as-a-Service paradigms, anyone in charge of their company’s infrastructure faces a difficult decision: should you build or acquire your own data center, lease space in a co-located facility, or rent a piece of the cloud?
It’s a question without a simple answer and one that will likely evolve over time. Factors complicating this calculation include business strategy, availability of resources, compliance, control, and scale. Where is the best place to spend your current capital? How quickly and how large do you intend to scale? What regulatory and compliance standards does your business need to adhere to? What sort of control systems do you or will you have in place at your site?
These questions and more must be considered when determining TCO, but no question is bigger than: “What cloud presence equates to a specific owned infrastructure footprint?” Comparing the cloud’s advantages in elastic computing and storage, availability, software tools, disaster recovery, and included labor, versus the owned infrastructure’s advantages in control, architecture optimization, and the ability to scale, is an incredibly difficult but necessary exercise in normalization that’s best performed with the application owner since no one better understands the needs of the infrastructure and business.
We started Coolan because we’re data-driven infrastructure veterans who’ve had to answer these tough questions a number of times in our collective past. This led us to create a TCO model to analyze the costs of your infrastructure options. The model provides a sound framework around which you can base strategic infrastructure decisions in light of TCO.
Our model compares three principal deployment scenarios: in the cloud, leased data centers, and acquired or custom-built data centers.
Before you run the TCO model however, it’s assumed the user has successfully identified the cloud presence and owned infrastructure footprints to be analyzed. From there, the model requires a number of inputs that cover CapEx (such as hardware development and procurement) and OpEx costs (such as staffing and power expenditures) for the cloud, servers, storage, and networks, as well as co-located (colo) and custom facilities. Results are adjusted for value/time concerns and presented as costs in today’s dollars. Finally, data driven decisions can be made.
Selecting an infrastructure option is very difficult. However, once you’ve done your homework, you can use our model to determine which scenario has the lowest TCO and is most appropriate for your business. Of course, specific needs of your application always take precedence and have the power to upset the results of any TCO analysis. But answering the big question of rent versus buy doesn’t have to be an intractable problem.
Take, for example, the effect of power costs between a colo and a custom-built data center. It’s well known within the industry that data centers have voracious appetites when it comes to consuming electricity, and there is an urgent need to rein in such consumption.
The following chart tells a tale of energy costs for both colo and custom-built data centers. For purposes of comparison, the model uses 4,000 compute and 625 storage servers for both scenarios, equating to about 1.5MW, which is a reasonably-sized colo deployment.
Here, the TCO model compares the ratio of energy cost in an 18-year TCO (in NPV/net present value) to energy cost. Notice that the cost of energy in the colo solution is higher than in the custom-built data center. But what’s surprising is that even at $0.50/kWh, the cost of energy is still relatively small compared to total cost in both cases.
Our model compared efficient facilities with an average PUE of 1.3 for the colo and 1.15 for the custom data center. Want to know what happens when the facilities are inefficient? Plug the numbers in the model to find out.
Shifting your company’s infrastructure requires considering all the pros and cons and is not a simple analysis. It’s easy to say that everything is headed to the cloud, but depending on your business objectives and the type of organization you’re running, the “old-fashioned” way of deploying in a colo space or building your own data center might still make sense.
We’ve developed one way for you find out: this TCO model is open and community-driven, so you can modify it to suit your purposes. We welcome you to share your changes and anything you discover during the process. Check out the model here.
We have pre-populated the TCO model with typical cost estimates for data center equipment, labor, etc., to indicate how the spreadsheet should be used and to give you an idea of how a cost comparison might shake out among the pre-determined deployment options (cloud, colocation, owned data center). The figures are based on 2013 market data and are provided as simplified representations. (Please check the “References” tab for more information.) We encourage you to click and play around with the spreadsheet so that it’s customized for your needs. A blank spreadsheet can be found here. To hear more about the TCO model, check out Coolan’s presentation at the 2015 Open Compute Summit.
