Determine True
Total Cost of Ownership

Debunking the Myth

VMware meets all essential customer requirements when companies virtualize their datacenters.

• Built on the robust, proven foundation of VMware ESX/ESXi
• Delivers a platform for shared IT services with VMware Infrastructure
• Provides a complete solution for virtualization management with VMware vCenter Server and the Application and Infrastructure Management suite
• Integrates with a customer’s entire IT infrastructure with broad software and hardware ecosystem support
• Has a proven success record with over 120,000 VMware customers and counting

But what about cost? Other vendors would like you to believe that VMware is too expensive. In fact, they commonly claim that VMware is three to five times more expensive than their own offerings. They base their claims on comparisons of upfront licensing costs.

Instead, comparisons need to be based on total cost of ownership (TCO). Looking beyond just the upfront license costs, any company doing a TCO analysis for virtualization must include the following in its calculations:

• Virtual Machine Density per Physical Server—How many virtual machines can run per host and therefore how many servers and software licenses do you need to buy?
• Operational Costs Savings—How does each solution improve your IT staff efficiency and reduce operational costs given how IT administration and maintenance costs dominate IT budgets today?

Maximize Virtual Machine Density per Physical Server

Before virtualization, IT would run one application per physical server so cost-per-server was a quick way to compare costs—it was a one-to-one relationship.

But once you virtualize, many applications (each in its own virtual machine) run on each physical server—it is now a many-to-one relationship. So cost-per-server comparisons no longer make sense. A much more accurate metric is cost-per-application because you want to know how much it costs to run the entire set of applications required to maintain business operations. To illustrate with an analogy, it is like asking: “Which is more cost-effective, a 4-door sedan or a 50-passenger bus?” The sedan may cost less upfront, but if your requirement is to transport a football team, then the 50-passenger bus is clearly more cost-effective! The cost-per-passenger is much lower because the bus has a higher passenger-per-vehicle density. Density matters in a many-to-one relationship.

Virtual machine density per host (number of concurrent VMs that can run on a physical server) directly affects cost-per-application. No other virtualization platform achieves the high virtual machine density of VMware Infrastructure / ESX and can still maintain consistent, high application performance across all running virtual machines.

VMware has invested in technologies to achieve very high VM density on ESX.

• Our advanced memory management features such as memory overcommit (with memory ballooning) and transparent page sharing, utilizes physical memory far more efficiently than any other virtualization platform on the market. In fact, no other virtualization platform has memory overcommit and transparent page sharing today, which severely limits the number of concurrent VMs other platforms can run on a physical server—customers end up having to buy more physical servers and/or memory. VMware ESX can commonly run twice the number of VMs on a physical server as our competitors. No other virtualization platform has these capabilities, meaning, fewer VMs per host, diminished performance, and larger hardware requirements.
• Scalable performance: VMware's engineers have fine-tuned VMware ESX/ESXi to achieve consistent, high performance when running many concurrent VMs on a physical server. The VMware direct driver model and virtualization-specific scheduler are key enablers of our more scalable performance. Others will focus on single VM performance running on a single physical server, but that does not tell the whole story. Single VM performance does not reflect what happens in the real-world with virtualization.

Being able to run your applications on fewer physical servers directly affects your bottom line by dramatically reducing hardware, software, power, cooling, and datacenter space costs.

Customers tell us that VMware DRS lets them run even more VMs per host (compared to running without DRS). They are willing to run at higher average server utilization levels with DRS because they view DRS as a “safety net”. Should a spike occur on any given server, VMware DRS will balance workloads within the cluster so that no virtual machine is starved of resources. Without DRS, customers estimate they would run at least 25-30% lower utilization levels to provide extra headroom. This translates directly into 25-30% more server hardware and software costs.

See how VM density affects the price per VM of VMware’s products and other virtualization offerings.

To top

Some have argued that they can add more memory to the Hyper-V or Xen host to achieve the same number of running VMs on VMware ESX. Even in this case, VMware still comes out less expensive or comparable.

Save on Operational Costs

IT management / operational costs can be several times greater than hardware and software acquisition costs over the lifetime of a server and must be factored into any total cost of ownership analysis.

You can directly reduce your operational costs by using the dynamic IT services built into VMware Infrastructure 3 that most other competitors do not offer. For example:

VMware VMotion enables planned server maintenance with no downtime impact on end-users. IT admins no longer need to come in on weekends or evenings (overtime pay) and spend hours contacting application owners to schedule a maintenance window. In a 150-VM VMware environment, a company can save an estimated $52,800 in IT administrative costs each year by using VMotion instead of scheduling downtime during evenings and weekends.

VMware Storage VMotion enables storage array maintenance and upgrades with no downtime impact to end-users. In a 150-VM environment VMware environment with 7.5TB of shared storage, a company can save an estimated $52,250 in IT administrative costs each year when performing storage array maintenance and upgrades.

VMware DRS saves IT from having to manually monitor VMs and manually move them to ensure proper resource reallocation. In a 150-VM environment VMware environment, a company can save an estimated $46,800 in IT administrative costs each year using DRS instead of manually monitoring workload and responding to customer calls when there are issues.

VMware HA automatically restarts virtual machines when hosts or individual virtual machines unexpectedly fail (unplanned downtime). This capability dramatically reduces the costs of lost end-user productivity due to the downtime. In a 150-VM environment, a company can save an estimated $60,000 in lost productivity.

VMware vCenter Update Manager automates scanning, tracking, applying, and remediating patches for the virtualization layer and guest operating systems. In a 150-VM environment, a company can save an estimated $149,000 in operational costs compared to applying patches manually. This figure does not even include the cost savings of using VMware VMotion with VMware vCenter Update Manager to patch the virtualization layer without taking applications down.

These savings would be lost by going with another solution that does not offer these dynamic IT capabilities.

Additionally, VMware also recently released its Application and Infrastructure Management suite of products—VMware vCenter Lab Manager, VMware vCenter Site Recovery Manager, VMware vCenter Stage Manager, and VMware vCenter Lifecycle Manager—targeted at automating management tasks that have traditionally been very time-consuming. Learn more about this suite of products in the Complete Virtualization Management section.

To top

Next Steps

Explore our Competitive Resource Center with virtualization whitepapers, reports, and videos or get a free copy of VMware ESXi right now to see how easy it is to create virtual machines, run multiple operating systems on a single server, and run even the most resource intensive applications on virtual machines.