When selecting a dual-socket blade server, performance per watt is an essential metric to measure performance. This white paper reviews results from a study that tested systems by running a variety of different blade configurations on each server. Using an industry standard benchmark, SPECjbb2005, the servers' Java performance was tested on systems from Dell, HP, and IBM. This white paper offers in-depth information on the following:
- System results when running 1, 2, 10, and the maximum configuration of blades
- A comparison of idle power usage results
- Important considerations for choosing blade servers for your data center