Capacity on Demand provides the ability to activate processors and memory units on your server when your workload requires the additional resources. To determine when to activate additional processors or memory units and how many new resources you require, monitor your trends in CPU and memory utilization by using a performance tool. Several performance tools are available to report CPU utilization information.
When computing average usage of all available processors, the system functions that report CPU utilization do not include the inactive processors in the total amount of CPU capacity. The inactive processors are not considered active within the various system functions that report CPU utilization percentages. The percentage of used CPU capacity is a calculated metric based on the amount of time that the processor was active within an elapsed time. This capacity is typically reported as a percentage, where 100% indicates that the processor was busy for the entire elapsed time. When multiple processors are present, CPU time must be adjusted to represent the average usage of all processors, so utilization is always reported as the percentage of total available capacity.