Thank you Stefan for developing the benchmark framework: https://github.com/boostorg/ublas/pull/57
I think it is a very good starting point to simplify the process of profiling.

If we integrate the tensor framework and start to compare with other non-boost frameworks we might need to change some functionality.
I would like to propose also my benchmark framework as an extension that also works for tensors. For an example, please see:
https://gist.github.com/bassoy/96b03bd9204758a3f4763ee8a212b3c1

Cheers
C