Servers are essentially large computers. However, the full extent of their capacity may be unused, largely because data center managers focus on reliability and not energy efficiency. Until recently, it was fairly common to have a separate physical server for each application. Additional servers were used for testing, staging, and in case of disaster, so there could be multiple servers per application (Energy Star, 2012). In 2012, the McKinsey and Company found the average server utilization rate to be 6% to 12%, and other studies have reached similar conclusions (Glanz, 2012).  

A virtualized server consolidates applications and sets of data, possibly from different customers' companies, so one server may now handle the tasks previously performed by multiple servers. Instead of running many servers with low utilization rates, you can run fewer servers with higher utilization rates. A New York Times article stated that one company has been able to operate at over 96% virtualization by queuing up large jobs and another company shrunk their server space by 60% after virtualization and server replacement (Glanz, 2012). The University of California in Santa Cruz had servers operating at a 5% utilization rate; using virtualization, these servers now operate at 70% utilization rate (LBNL, 2007, page 2).

In addition to a one-time or periodic virtualization, workloads can be orchestrated among servers to maximize the concentration of computing activity on active servers while maximizing the opportunity to power down idle servers for a longer time. Some strategies to accomplish this include: 

• MAID (massive array of idle disks) concentrates popular data on a new “cache” disk (Colarelli, 2013).
• PDC (popular data concentration) uses a subset of the server for the most popular data (Pinheiro, 2004).
• Power-aware caches are used to house the data of spun-down servers to increase their idle time (Francis, 2004).
• Write-offloading diverts write-access from spun-down disks to active disks, localizing write access (Narayanan, 2008).
• SRCMap (Sample-Replicate-Consolidate Mapping) is similar to MAID and PDC and uses write-offloading (Verma, 2010).

Servers that are good candidates for virtualization and have similar workload types are grouped and then moved to as many physical servers as are needed to accommodate them in virtual form. It is important to observe server loading for a month to see when computing activity is busiest and combine servers with complementary workloads (e.g., one is busy during the day and another is busy at night) (Energy Star, 2012). Servers that are selected must have adequate memory for the task, as this can be more of a limiting factor than processor speed. An Information Week survey of large data centers found that the median memory (RAM) capacity was 48 GB per server and almost a fourth of these data centers used servers with 128 GB of memory (Marko, 2012). Consolidating servers also consolidates the cooling load. If the virtualization process continues, total data center use may become considerably higher. Therefore, check that the existing cooling system is up to the task (Energy Star, 2012).

Virtualization may be done in phases, as suggested by the Green Grid. The first phase is virtualizing non-critical workloads. The second phase includes critical workloads, which usually involves a need for additional storage and network reconfiguration. The third phase is implementing a “virtual first” policy for future servers. The final phase is to out-source computing to the cloud (Talabar, 2009).

As of 2011, 55 utilities in 20 states offered financial incentives for virtualization projects. Most of these provide a simple prescriptive incentive: $X per server removed. Calculated or customized incentives are more accurate but may be less applicable for small- and medium-sized businesses. Some utilities offer $/kWh for any documented data center savings. The Sacramento Municipal Utility District (SMUD), for example, offers $.04 per kWh saved up to 30% of the project cost or $150,000. Some utilities may want to consider a hybrid model that offers options of prescriptive, calculated, and customized incentives. Utilities should encourage recycling of old servers and work to ensure that those that are not recycled are at least removed from the grid. Utilities may want to skip inspections of smaller projects to help control administrative costs. Finally, larger utilities may do well to assign utility staff to learn about virtualization so they can advise customers.

A side benefit of virtualization incentive programs for utilities is they provide an opportunity to develop a relationship with IT managers who may be willing to consider infrastructure efficiency upgrades after the virtualization project (Lester, 2011). NRDC suggests that these companies’ use of virtualization can be expanded through utility programs with robust marketing and outreach, training, and education programs for IT managers of small server rooms and the IT service provider firms that support them. This would ideally be at a time when IT managers are preparing for a major server room upgrade anyway. Utilities can also articulate the non-energy benefits of virtualization that would apply to small businesses that might lack utility bill incentives. Utilities could also act as aggregators for manufacturers and providers of virtualization services that have not targeted the small business market (Bennett, 2012).

Because large data centers have more readily embraced virtualization than smaller data centers, utilities may be concerned about free-ridership in justifying their incentives to rate-payers. The Green Grid has found that, while many companies may have adopted virtualization, it is typically only at a shallow level with very small consolidation rates. Therefore, utilities still stand to gain substantial energy savings by incentivizing and encouraging customers to get beyond the “low hanging fruit” ("innovation" servers) into deeper savings ("mission-critical" servers) (Talabar, 2009).

Virtualization products and services are provided by VMware, Citrix, Oracle, Red Hat, and Microsoft with costs ranging from $200 (on sale) to tens of thousands of dollars.  Most require or at least encourage annual subscription and support fees. 

Over the past five years, virtualization has become fairly common among large, dedicated data centers, but is still less common among smaller server rooms in businesses. In a 2012 study, the NRDC reported that of 30 small businesses (ranging from 3 to 750 employees) surveyed, only a third used virtualization at all, only a quarter of their server stock has been virtualized, and less than a quarter of these businesses planned to increase their adoption of virtualization. NRDC writes, “Unless something is done to encourage small businesses to adopt virtualization, we may see virtualization penetration rates continue to remain low in that segment" (Energy Star, 2012) (Bennett, 2012 Pg 3-5).

Low adoption rates of virtualization is primarily due to unaligned incentives. Over half of the small businesses have leases with a fixed fee for utilities and, therefore, do not stand to benefit directly from cutting energy use. Furthermore, over 90% of small businesses lack the ability to determine server room electrical energy use even if they do pay energy bills directly (Bennett, 2012 Pg 6). However, server rooms can account for 30% to 70% of these organizations' energy costs. Half of these small businesses plan to upgrade their server rooms in the coming year, but with these disconnects and barriers to encourage efficiency improvement projects, these projects are unlikely. 

Of those IT managers who are interested in improving energy efficiency, many have trouble convincing their managers that it is worth their time and effort to learn about the new technologies and strategies that could help them achieve those goals. The IT managers also have trouble reassuring their management about data security concerns. A smaller subset of these business use cloud services for off-site computing or rent their servers so they are not in a position to implement virtualization (Talabar, 2009).

Server virtualization has been discussed at numerous conferences over the past 5 to 10 years. Most large data centers have adopted this strategy to some degree. Software firewalls and data management tools required to accomplish this safely have been developed and consulting firms can help with implementation.

According to InfoWorld, sales of virtual machine and cloud system software grew 17.8% in the first half of 2013. "Virtual machine software unit shipments still remain healthy and growing, but have seen some slowdown in mature markets that have high virtualization rates," IDC analyst Gary Chen said. "Business models are shifting as well, with the hypervisor drawing less direct revenue and increasingly becoming an embedded feature of operating systems and cloud system software." Other factors impacting the virtualization market include stiffer competition for VMware from companies such as Red Hat and Microsoft, as well as the rise of open-source cloud software such as OpenStack, according to Chen (Kanaracus, 2012).

Non-energy benefits include reducing the number of servers needed, thus reducing the resources needed to purchase, set up, and maintain the servers, as well as reducing the number of software licenses needed. 

Virtualization improves a data center’s availability, flexibility, and disaster recovery ability by allowing virtual servers to quickly be moved from one physical server to another, even from one site to another (Bailey, 2009 Pg 5). It also helps that virtual servers can restart applications after a disaster much faster than physical servers (Energy Star, 2012). Virtualization allows data centers to be more responsive to customers’ needs by reducing the time required to deploy a new server (hours instead of days or even weeks required with a physical server deployment) (Bailey, 2009 Pg 5). Virtualization allows data center operators to move entire systems from one physical server to another within seconds to optimize workloads and to allow them to perform maintenance without disrupting work processing. 

The Uptime Institute estimates that decommissioning a single rack of servers generates annual savings of $500 in energy, $500 in operating licenses, and $1,500 in hardware maintenance costs (Energy Star, 2012) (Mstanberry, 2012). VMware estimates that virtualization can cut capital expenditures by 60% and operations expenditures by 30% while providing added monitoring and analytical capability for improved troubleshooting and performance (VMware, 2013). Virtualization is also excellent preparation for transferring computer services off-site to a cloud service.

Virtualization can also reduce real estate costs. The University of California in Santa Cruz had 188 physical servers and was nearly at their maximum capacity. Since virtualization, their number of servers increased to 240, but they still have available capacity and did not need a bigger building (LBNL, 2007 Pg 1).

Southwest Illinois College performed an economic analysis of upgrading their 35 servers in 2011, both without virtualization and using virtualization to consolidate to four physical servers. This generated $280,000 in savings over three years; $150,000 of that was for direct savings on server hardware and networking minus the cost of virtualization software and $130,000 was for indirect savings on cooling costs, server provisioning, and procurement.

IDC studied virtualization at a Landmark Healthcare facility in 2009. There, 63 servers were consolidated onto just three physical servers. Virtualization software cost $3,600 but they saved $60,000 annually for server hardware, backup system software, and operating system licenses (Energy Star, 2012 Pg 16).

According to Gartner, a multi-national information technology research and advisory company, 30% to 60% of virtualized servers installed between 2010 and 2015 may be less secure than the physical servers they replace. Virtualization is not inherently insecure, but it needs to be done properly to maintain security, and that process is improving. Gartner has identified six contributing factors to this rising vulnerability. One is not involving the information security team in the initial architecture and planning stages.  Another is the possible risks of the hypervisor and virtual machine monitor from hackers. To address this, virtualization vendors should provide a means for detecting when the hypervisor/VMM (virtual machine monitors) layer has been compromised (GlanzGartner, 2012). 

Obtaining per­mission to virtualize servers may be challeng­ing. Ownership of servers may be unclear, and efforts to locate the appropriate parties can delay a virtualization project. Internally, the relationship between IT staff and building management staff may need to change to take full advantage of virtualization opportunities (LBNL, 2007 Pg 2). According to the Uptime Institute, only 20% of data center departments pay their own power bill. For the rest, their facilities or corporate offices take care of it, so IT managers are less aware of the impact of their actions on energy use and may have less incentive to work hard to reduce energy use (Mstanberry, 2012 Pg 2). 

The primary competing technology (or strategy) for virtualization is out-sourcing IT services to huge data centers operating in the cloud. While these large data centers are likely much more energy efficient than small server rooms, many small businesses are uncomfortable with cloud computing and do not trust cloud services with their private information (NRELBennett, 2011). Virtualization will help small businesses prepare for transitioning to cloud computing in the future. 

The other competing strategy is avoiding virtualization, avoiding the costs and complexities, and maximizing data security.