Performance case

MIPS and MSU ratings account for all active processors on a particular hardware configuration, yet the Command Server or a single transformation can only use a single processor at any given time. Therefore, comparing MIPS or MSU ratings between machines to estimate transformation speed will not produce reliable estimates if the number of processors differs. To derive MSU ratings that can provide valid estimates, the number of processors must be taken into account.

Here is a hypothetical case using fictitious configurations to illustrate how estimating performance results might be affected by various factors. The considered factors include hardware configurations, MSU ratings, and taking into account the number of processors on the machine (normalizing) used to run maps or perform transformations.