It's hardware-level - division requires more CPU cycles than multiplication on most processor architectures, making this optimization pattern relevant across virtually all programming languages.