The CompaqNonStop Himalaya server excelled in an international Standish Group independent study of enterprise servers and their Total Cost of Ownership (TCO). The study scrutinised servers from top-name vendors including HP, IBM and Sun Microsystems and was based on more than 150 interviews with companies the majority of which had revenues exceeding $500M.
Lowest Downtime, Lowest Application Cost, Lowest Application Downtime Cost The study found that the Compaq NonStop Himalaya server not only had the lowest system downtime, but also the lowest application cost and the lowest application downtime cost. The superiority of the Compaq NonStop Himalaya server becomes more marked as the cost of the average lost transaction increases. In a scenario where the average cost of a lost transaction was $50 3/4 a realistic price in stock trading environments 3/4 the average cost of downtime per transaction on the Compaq NonStop^O Himalaya server was more than 70 percent lower than that of its nearest rival.
The study cites three characteristics of the Compaq NonStop^O Himalaya server which contribute to its superior TCO performance: transaction-based analysis, continuous availability design and having many applications that require few people to manage. The other servers under examination were HP Clusters, IBM Parallel Sysplex, Sun Clusters and Compaq`s TruClusters . The Standish Group used its Standish Transaction Model to evaluate cost performance 3/4 a model it has used for years and one found to be `highly accurate`.
Nonstop^O Himalaya Platform 3/4 the Engine for Large-Scale, Internet-driven e-business Environments
The Standish findings underline the business case for the Compaq NonStop^O Himalaya server as a core transaction engine for large-scale, Internet-driven e-business environments where reliability and scalability are key. "People know that achieving 24x7 is difficult. IT shops are spending huge amounts of money attempting to achieve 24x7, and they are often failing to achieve it. The Standish report highlights the fact that NonStop^O Himalaya servers provide the very highest levels of availability at the lowest cost," said Mike Clark, Manager - Tandem Division at Compaq Computer. "Any Chief Information Officer contemplating purchases to gear up for the high-speed, Internet-driven market cannot now afford to ignore Compaq NonStop^O Himalaya servers to support solutions which stay up continuously, and which need to grow quickly and massively".
Most Reliable TCO Study Design To Date
Presenting its findings as the `most reliable TCO study design to date`, the Standish Group describes the study as `a true apples to apples comparison` and one closely based on reality: "This study is real life. This is not a made-up bench-mark where engineers play their sinister games to get the last ounce of performance. These study results reflect real people, operating real mission-critical applications, with the tools they have to work with, to the best of their abilities."
Scope and Accuracy of the Study
The main purpose of the TCO research was to evaluate the operational and cost environment of the various systems. It looked at system cost, most mission-critical application costs, utilisation, down-time and quality of service, as well as hardware cost, software cost, software and hardware maintenance, manpower for both system and the most mission-critical application, and other costs including such things as space, paper and electricity.
The accurate calculation of TCO necessarily involves a variety of often complex factors. Previous analyses have sometimes over-emphasised the significance of list price while omitting the critically important data associated with applications usage. To compare one system to another, Standish has developed the Standish Transaction Per Second, a standard measurement which when applied, enables an `apples to apples` cost calculation. The Standish TCO model is the "most germane for evaluating enterprise servers, while other models are best suited for desktops and PC servers."
Share