I have discussed this with @jorisv and @fabinsch a bit.
The idea is that we can use a more advanced code profiler such as tracy (which requires instrumenting the code) to better assess performance bottlenecks in some critical (sub-millisecond) pieces of code with high accuracy.
I have discussed this with @jorisv and @fabinsch a bit.
The idea is that we can use a more advanced code profiler such as tracy (which requires instrumenting the code) to better assess performance bottlenecks in some critical (sub-millisecond) pieces of code with high accuracy.