evaluating how well various computers will handle an organization's workload by running a representative set of real jobs on the different machines is called