Hi. Welcome to Ingram Micro.

Please choose your role, so we can direct you to what you’re looking for.

If you’d like to learn more about Ingram Micro global initiatives and operations, visit ingrammicro.com.

How Data Center Design Impacts Efficiency and Profitability

July 28, 2017

The cost of data center operations can give CIOs sticker shock. When you assess the total cost of ownership (TCO), you have to consider (1) capital expenditures (CAPEX) such as cost to design, cost to build, and commissions and (2) operating expenses (OPEX) such as maintenance, staffing, upgrades, swap-outs, etc. Those who have done the data center CAPEX/OPEX analysis know that typically OPEX will exceed CAPEX. In order to get the most out of data center operations, reduce OPEX, and increase the data center’s contribution to the bottom line, you have to examine all facets of data center design closely.

The cost of running a large data center can run from $10 million to $25 million annually. Typically, 42 percent of costs go to hardware, software, disaster recovery, and networking costs. The remaining 58 percent goes to building infrastructure such as cooling, taxes, and labor. In fact, up to 40 percent of overall costs can go to labor alone.

As you can see, when it comes to managing the efficiency and profitability of a data center, designing for optimal efficiency matters.

Controlling Cooling and Energy OPEX Costs

Energy and infrastructure-building costs are going to be among the biggest operating expenses, especially electricity and cooling. According to American Power Conversion (APC), 20 percent of data center operational costs go to electricity; 18 percent to power and service equipment; and 6 percent to heating, ventilating, and air conditioning. If you can design a greener data center that requires less power and less cooling, you can substantially lower TCO.

The cost of cooling data centers has already started dropping, and it is no longer necessary to maintain a constant temperature between 64 and 68 degrees, and you do not have to scale cooling as you add new hardware. Today’s data center equipment is built to operate at higher temperatures, so less cooling is required. In addition, most data center hardware is now equipped with internal intelligence for thermal and power monitoring. If you use the thermal intelligence in the hardware, you can isolate hot spots. Heat monitoring makes it easier to plan for peak operational periods and manage cooling as needed.

Middleware is also available to monitor data center temperature and power consumption. Data center infrastructure management (DCIM) systems can monitor and manage data center temperature and power in real time. Using software to monitor and log data center performance also makes it easier to improve design. For example, minimizing the number of idle services can reduce energy and cooling requirements by as much as 15 percent.

In some cases, you can locate the data center in markets where energy is less expensive. In order to offset cooling and energy costs, many data center designers also are working with local climate. More data centers in high-desert or mountain locations use cool, dry external air to offset AC cooling.

Reducing IT Load

As part of data center efficiency, it’s always best to try to optimize hardware and minimize the number of redundant systems. One of the best ways to eliminate redundancy is with virtualization.

Using virtualization, you can run multiple applications in the same machine, which eliminates unused servers. In many data centers, 10 percent of servers are never used at all. With virtualization, you can increase server efficiency by 75 percent, which substantially reduces OPEX and CAPEX.

Even with server virtualization, one-third of computing costs are going to data storage. You can reduce the cost of storage through consolidation and virtualization. Determine how much storage you need on site versus in the cloud. If you can consolidate and reduce your data storage consumption, you also reduce both CAPEX and OPEX.

Offloading Costs to the Cloud

Migrating data center resources to the cloud saves on both CAPEX and OPEX. If you use cloud-based computing and data storage, them you adopt a pay-as-you-go model for data center resources. This eliminates server and storage redundancy as well as operating overhead for cooling, power, maintenance, etc.

Cloud data centers are designed differently from dedicated data centers. They are built to scale differently and are not constrained by the same limitations as a dedicated data center. That also means the economics are very different. For example:

  • Labor makes up only 6 percent of overall operating costs.
  • Power distribution and cooling are only 20 percent.
  • Computing costs are 48 percent.

And for customers, it means sharing operating costs so that the overall costs can be substantially lower. And the cloud supports extensibility and elasticity without CAPEX; you can add computing and storage on demand.

Shorten Time to Data Center Profits

Of course, cloud computing is not the answer for every organization. Some organizations need to consolidate control and manage security for regulatory compliance or strict security, or they have other reasons to develop their own data center without cloud support. Designing and building a green, stand-alone data center that makes optimal use of power, cooling, virtualization, and other best practices can reduce operating costs and lower TCO. There are also strategies that can help shorten time to data center profitability.

The right reseller partner with data center consulting expertise can recommend ways to save on hardware in order to cut CAPEX, for example. Skilled resellers and data center architects can determine where to get top-grade hardware for less and put it to maximum use. Server remarketing programs also can obviate major capital outlays. By reselling outdated servers, data center customers can stay current with the latest technology at lower costs.

Of course, getting assistance from experienced data center architects can be the best strategy to reduce data center costs and shorten time to profitability. The right expert can identify what resources are needed, help spec the data center for future needs, and blend in-house and cloud computing resources for maximum efficiency. Partnering with experienced partners like Ingram Micro can help you help your customers make the most of their data center investment.