[FEA]: Analyze histogram/even benchmark performance

[danielfrg]

Is this a duplicate?

• I confirmed there appear to be no duplicate issues for this request and that I agree to the Code of Conduct

Area

cuda.compute (Python)

Is your feature request related to a problem? Please describe.

Follow up from the new nvbench Python benchmarks and comparison with C++ ones: #7341

We are getting more than C++ performance! 107% (aggregated numbers).
So we should look more into that.

Describe the solution you'd like

Mabye something is not fully matching on the python side.

Describe alternatives you've considered

No response

Additional context

No response