I forgot that I used to be a MMX and 0x86 assembly optimisation specialist when I was young. Today I remembered, and was surprised to find that there are still traces of it on the Internet. Here are some Pentium optimizations in the source code of the Raine emulator, when performance used to matter a lot. https://github.com/zelurker/raine/blob/a8596c6a5e71c3982c3aaf6896b3c857191a5012/source/video/i386/spr8x8_8.s#L44-L48