Lowering cost with capacity on demand

Finding the right balance between what you need today and expect to require tomorrow can be a difficult task. Most user firms either overshoot their requirements or fall short of what they need. Puneet Gupta says the solution is capacity on demand

A prerequisite of any IT purchase today is the ability of the IT team to demonstrate to the executive team the expected return on investment (RoI), return on assets (ROA) and impact on total cost of ownership (TCO). The reality today is that instead of increasing IT budgets to meet increasing and often unpredictable requirements, budgets have actually stagnated or declined from previous years.

A recent ITG Group study showed that the combined spending on hardware, software, staff, facilities and services for servers alone was expected to account for approximately 40 percent of corporate IT expenditure during 2001, compared to 31 percent in 1996. With server infrastructure costs growing more rapidly than any other IT resource, this area is receiving special attention as one where a dramatic change in modus operandi may be required to reconcile growing processing needs with declining IT spend.

The focus on getting tangible returns out of an IT investment is not a new one, and TCO analysis especially is receiving a lot of attention from customers and server hardware vendors alike. The current dilemma facing enterprises today is this: how do they to cope with a dynamic market environment with its multiple peaks and troughs while keeping TCO as low as possible.

Enter Capacity on demand

Capacity on demand (COD) is a relatively new phenomenon in the IT industry. It allows companies to receive equipment with more computer processing, storage, or other capacity than the company needs at the time of purchase, and keep that extra capacity remain unused and unpaid for until the company actually requires it.

The benefits are obvious. The road to a major IT purchase is often a long drawn-out affair, complete with detailed tender documents and drawn-out evaluations before the contract is finally awarded and implementation can begin. Sometimes the process takes so long that the business requirements at the time of the implementation already exceed the capacity of the system. At other times, due to overly confident estimates, the enterprise ends up with excess capacity it may not need for years.

The beauty of COD is the fact that the enterprises only pay for the additional capacity as and when they need it. And because it is already there, the new computing resources can be activated almost instantaneously, simply by obtaining approval codes from the vendor.

COD also helps enterprises avoid the current extremely wasteful trend of server under-utilisation that’s prevalent in many organisations. In a recent case study conducted by IBM, one enterprise—that had 450 Windows servers, 200 UNIX servers and a couple of mainframes—had server utilisation rates as low as 5 percent for its Windows servers, and slightly better at 12 percent for its UNIX servers. The mainframe average utilisation was much better at 62 percent. What is worrying about these statistics is that this is a fairly typical scenario—but one that COD is perfectly designed to address.

Why pay now when you can pay later?

No business should be a victim of its own success. Should business transaction volumes increase, they must be able to smoothly scale up operations without a hiccup. COD enables them to build scalability into their IT infrastructures and processes from the start, reducing the risks and costs of having to add new and untested hardware in a hurry.

Vendors are already offering COD features such as ‘capacity upgrade on demand’ and ‘on/off capacity on demand,’ both of which allow companies the flexibility of getting more processing power when they need it, and more quickly. For example, with ‘on/off capacity on demand,’ a company can turn processors on when it needs them—and turn them off when it doesn’t. The company pays only for what it uses and not a penny more. Companies will find these features cost-effective and timesaving alternatives to more traditional and time-consuming methods of upgradation. ‘On/off capacity on demand’ makes it easier for enterprises to scale or adjust their IT infrastructure on-the-fly.

In a typical example of ‘capacity upgrade on demand’ (CUoD), an enterprise may receive from their vendor a fully configured 24-processor computer server but only have to pay for a 16-processor server initially as that is all they need at the time of purchase. The CUoD operating environment provides the flexibility (and speed) for the enterprise to turn on the extra capacity when it needs it. In this scenario, they would simply contact the vendor to order an activation code to permanently turn on the extra processors, and the vendor would bill them accordingly.

The result? Capacity is added in a timely, cost-efficient and least disruptive manner, enhancing operational effectiveness, customer service and employee productivity.

But why stop here? In the bid to make it even easier for enterprises to adjust their IT infrastructure on-the-fly, two new offerings have recently hit the market.

• ‘Memory on Demand,’ which allows customers to install features with unused memory capacity that can be activated in 4 GB increments
• ‘Standby storage capacity on demand,’ which enables customers to acquire terabytes of storage capacity quickly to meet growing data requirements. Coupled with storage consolidation and networking solutions, storage on demand technologies will help storage administrators improve utilisation rates while easing management woes. With information becoming the most important corporate asset and one study showing that about 250 MB of data is being collected per-person, per-year, data growth is outstripping storage capacity in many organisations. The adoption of on demand capabilities makes perfect sense in the storage arena at this time.

Conclusion

Installing COD-enabled servers and storage enables enterprises to prepare for increases in demand without up-front capital outlay and keeps utilisation rates at optimal levels. It saves time and effort by avoiding multiple, lengthy approval cycles and time-consuming paperwork. The efficiency of IT investments is improved by allowing organisations to acquire server and storage capacity in advance and pay for it only when they use it.

But most importantly, it helps enterprises align the costs of doing business with revenue growth, improving RoI and ROA. In an e-business on demand world, capacity on demand makes eminent sense.

The author is country manager–eServer pSeries, IBM India