This has a chapter on low level optimisation in Owl: https://ocaml.xyz/book/core-opt.html, which includes how core functions are implemented in C, and how OpenMP is utilised etc. Besides, Owl relies on certain libraries…
A trivial and tiny little bit improvement: `objdump -d /usr/bin/* | cut -f3 | grep -oE "^[a-z]+" | sort | uniq -c | sort -k -1rn | head -n5` to get the top-5 instructions.
A similar list: https://prism-break.org
This has a chapter on low level optimisation in Owl: https://ocaml.xyz/book/core-opt.html, which includes how core functions are implemented in C, and how OpenMP is utilised etc. Besides, Owl relies on certain libraries…
A trivial and tiny little bit improvement: `objdump -d /usr/bin/* | cut -f3 | grep -oE "^[a-z]+" | sort | uniq -c | sort -k -1rn | head -n5` to get the top-5 instructions.
A similar list: https://prism-break.org