Recently vendors have been pushing computing on demand. You've all heard the pitches: "Buy only what you need", "Pay only for what you use", etc. In this time of tight budgets it sounds good. It should be a big success. Or should it?
A look back at history suggests that computing on demand may not be the long term success everyone is predicting. Back in the 1960's and early 1970's IBM rented machines. Remember the hour meter on the old S/360 and S/370 machines? Every month your IBM customer engineer dropped by and recorded the hours used. Then IBM sent you a bill for the usage. If required, IBM put in a faster machine and you used whatever you needed. Sounds a whole lot like computing on demand.
Customers (you and I) figured out we were paying a premium price for this service. The accounting department didn't like the bills that varied every month; accountants prefer fixed costs that are easily predictable. In response to fiscal pressure, the market changed to either a fixed lease regardless of usage or outright purchase of computing power. This led to a whole field of computing known as performance management. The company that figured out how to get more use out the hardware had a lower cost per unit and a competitive advantage.
Now we are back to variable costing plans. In times of contraction, this is good because computing costs can decrease. As soon as expansion starts again, the variable costs will go up. It won't be long before someone figures out fixed cost may be better in an expanding economy.
There are signs this is already happening. IBM had to provide guarantees to z/OS customers that their measured usage charges would not exceed their fixed software maintenance costs for a period of time, usually between one to two years. Also, a couple of outsourcing deals where the cost was based on the number of transactions or amount of merchandise sold have recently been canceled or re-negotiated. A sign of things to come?
Bottom line: computing on demand may be of benefit for companies that have highly variable workloads but not for fixed demand organizations. You need to understand your workload and match it to the most cost effective solution.
About John Weinhoeft:
For the past 30-plus years John Weinhoeft has had his hand in the computer industry. He recently retired from designing and managing the State of Illinois' centralized computer systems that served 100 agencies. John has authored and edited a number of analytical books published by Computer Technology Research Corporation. He is, or has been, a member of several computer organizations including the Computer Measurement Group and Central Illinois Personal Computer Users Group.
Send John your comments