|
For organizations that are weighing the choice between deploying a large number of servers, total cost of ownership (TCO) analyses favor blade server systems over rack-optimized server systems for reductions in both capital and operating expenses.
Blade server systems are all about exploiting the economies of scale when deploying servers in volume. Analyzing data provided by HP, IDC validated the nature of the expected cost savings:
From a capital expenditure point of view, hundreds of independent rack servers require thousands of repetitive parts. This need is apparent when integrating servers with local area networks (LANs) and particularly with storage area networks (SANs). Without a shared backplane, each standalone rack server requires adapters, cables, and supporting switches, and there is no economy of scale as the number of servers grows. The blade server system, with its integrated backplane, consolidates LAN and SAN access and thus requires far fewer interconnecting cables and devices.
From an operating expenditure point of view, further economies of scale emerge. Reductions in power, cooling, and datacenter space accumulate to lower facilities cost by an estimated 25%. Most importantly, the cost of initial system deployment drops dramatically as the software combines with the blade server system to automate system installation tasks. Moreover, consolidated networking for LANs and SANs reduces the number of cable pulls. Thus, installation tasks that once took a few hours per server can be accomplished in a few minutes per blade. In addition, costly tasks such as running hundreds of cables disappear entirely.
In this context, IDC believes that the HP BladeSystem c-Class is an important new platform for IT organizations to evaluate when weighing the relative TCO associated with building out server and infrastructure resources. With next-generation HP Thermal Logic technology to reduce energy consumption and improve cooling efficiency, the HP BladeSystem c-Class helps to alleviate power consumption limitations that are emerging in some datacenters. HP Virtual Connect architecture streamlines LAN and SAN connectivity and expedites the provisioning process.
Built-in HP Insight Control management improves the ability of system administrators to deploy systems rapidly, catalog blade system resources, provision infrastructure, and monitor overall system health.
Situation Overview
In today's enterprise, control and reduction of expenses are top-of-mind issues. Both capital costs and operational costs are fundamental to any discussion about running and growing a business. Such costs extend to all lines of business within an organization as well as to the IT group.
However, when it comes to IT, identifying expenses may not always be as straightforward as with other business units. Primary expenses, such as the cost of hardware or software licenses, are relatively easy to see. However, secondary expenses, such as the hardware cabling costs, facilities charges, and personnel hours needed to configure the systems, are often not accounted for with precision. Tertiary expenses, which include the cost for service, maintenance, and retirement of old servers, are rarely allocated when estimating the TCO for a new IT investment.
The primary TCO metrics are relatively obvious and intuitive. They include up-front capital expenses that are accounted for when a company does a workup of the costs of deploying new IT hardware and services. Secondary TCO measures, which are primarily operating expenses, will be less apparent and will often require more research to quantify accurately. TCO analysis requires digging in to what system administrators do and how the datacenter is architected. Tertiary TCO expenses must be estimated by attempting to anticipate what future demands the installed hardware will place on the IT budget. These costs are all a part of the overall IT spend. The precise allocation of cost to inform investment decisions is the goal of TCO analysis.
Typically, IT customers will track the costs of infrastructure acquisition and operation over time so that they can make important decisions about capacity planning and computer acquisitions going forward. IT planners primarily use one-year, three-year, and five-year time horizons when conducting capacity planning for their computing systems.
How TCO Is Useful for IT Buildout
TCO analysis can be highly useful in relation to IT buildout in several ways. First and foremost, it can help customers identify the true cost of architectures for their datacenters before they invest in the infrastructure components.
|
