Running benchmarks in Rust is done via
cargo bench, a specific
cargo which is executing all the methods annotated with the
#[bench] attribute. This attribute is currently unstable and
available only for the nightly channel.
.NET users can make use of
BenchmarkDotNet library to benchmark methods and
track their performance. The equivalent of
BenchmarkDotNet is a crate named
As per its documentation,
Criterion collects and stores
statistical information from run to run and can automatically detect performance
regressions as well as measuring optimizations.
Criterion is possible to use the
#[bench] attribute without moving to
the nightly channel.
BenchmarkDotNet, it is also possible to integrate benchmark results with
the GitHub Action for Continuous Benchmarking.
in fact, supports multiple output formats, amongst which there is also the
bencher format, mimicking the nightly
libtest benchmarks and compatible with
the above mentioned action.