Need to make sure the benchmarks are run on the same machine both pre- and post-change to ensure the comparison is meaningful. Also need a way to expose this information in a meaningful way so that true performance regressions are properly highlighted.
Need to make sure the benchmarks are run on the same machine both pre- and post-change to ensure the comparison is meaningful. Also need a way to expose this information in a meaningful way so that true performance regressions are properly highlighted.