Microcode in Intel's 8087 floating-point chip: the scale instruction
An analysis of Intel's 8087 floating‑point coprocessor reveals the microcode behind its SCALE instruction, which multiplies a value by a power of two. The piece explains the encoding, execution steps, and performance impact, offering insight into early hardware design that shaped later x86 floating‑point operations. Hacker News discussion adds code snippets and notes modern emulators that replicate the behavior.