CT+ recommendations are based on profiling the energy consumption of Java collection types (see paper), collected by CT+ profiler. In order to replicate the results and also in order to look into new areas, e.g. JAX-RS applications (#1) or Java Streams, a benchmarking suite for measuring the energy profiles of different impletations needs to be created.
The Java Microbenchmark Harness (JMH) should provide a very good basis to recreate the test suite of CT+ and to use perf as a basis for measuring the energy consumption (on Intel Hardware perf uses Intel's RAPL interface, see here).
This should provide then a good basis to integrate this JAX RS benchmarking guide to tackle #1
CT+ recommendations are based on profiling the energy consumption of Java collection types (see paper), collected by CT+ profiler. In order to replicate the results and also in order to look into new areas, e.g. JAX-RS applications (#1) or Java Streams, a benchmarking suite for measuring the energy profiles of different impletations needs to be created.
The Java Microbenchmark Harness (JMH) should provide a very good basis to recreate the test suite of CT+ and to use
perfas a basis for measuring the energy consumption (on Intel Hardwareperfuses Intel's RAPL interface, see here).This should provide then a good basis to integrate this JAX RS benchmarking guide to tackle #1