In 2015, we converted MiDataSets and cDataSets to open Collective Knowledge Format (CK) and moved them here: http://github.com/ctuning/ctuning-datasets-min
CK allows you to simplify compilation and execution of these benchmarks across multiple platforms (Windows, Linux, MacOS, Android ...).
In this project, we collect multiple datasets for a MiBench benchmark (free, commercially representative embedded benchmark suite) to create the conditions for a more realistic benchmarking...