Chapter on low-level CPU optimizations

[doyougnu]

Things like:

• avoiding pipeline hazards
• fitting data structures and functions on a single cache line
• emitting and reading GHCs gas assembly
• Exploiting better branch prediction

The key aspect of these should be that the developer can read the chapter and use its techniques on their own code without being a GHC developer or a master of Cmm and GHC's RTS

Some resources:

• https://en.algorithmica.org/hpc/complexity/
• https://www.felixcloutier.com/x86/

Some memory footprints:

• https://blog.johantibell.com/2011/06/computing-size-of-hashmap.html
• https://wiki.haskell.org/GHC/Memory_Footprint