Here's my stab at doing this: https://gist.github.com/ddrown/3c6de60a ... 4ec350a9ca
I can't explain why my math is slower. Maybe the part of the code that I guessed at is different somehow?
Excellent! Yes, that is exactly what I was thinking (but too 'buzy' to do )
One of my concerns with any benchmarks is being able to reproduce the results; one of the key features of science.
Needed (at least):
- the entire source code,
- libraries (or at least library versions),
- compiler version and compiler flags,
- linker flags/link script,
- MCU and clock speed.
The difference may be compiler flags, compiler version, library (different micros?); even the link script can affect the results.