Chapter 14. Benchmarking
A benchmark is a standard against which things can be compared. The term is also used to refer to a standardized test or experiment with results, that when the test is applied to different systems, can be used to compare those systems.
In the development environment, these two meanings are practically inseparable. The benchmark results (i.e., the standard) can only be understood by understanding both the test that produced the results and the configuration of the system under test. Two sets of results from two different systems cannot be fairly compared unless the same test was used and the configurations of both systems are clearly understood. In fact, if there isn’t enough information for you to reproduce ...