Add benchmarking for significant operations

[TomAFrench]

This is a high importance library but we don't have any measurements on performance. I've described the CI setup necessary to do this in noir-lang/noir-library-starter#7.

@noir-lang/libraries can you add benchmarking for the gate costs of important functions so that we can track how these change as development continues.

[kashbrti]

I've been debating what good benchmarks for bignum would be.
Does reusing some of the test functions for benchmarking make sense?

[kashbrti]

this is being tackled in #100