Compilers on some processors will replace multiplications with a short sequence of shifts and adds, if possible. For example if variable Y is to be multiplied by $9,$ the replacement code could shift Y left by $3$ bits $($equivalent to multiplying by $8)$ and then add the original Y$.$ In this case, the combination of one shift and one add replaced the original multiply. $($The motivation is performance $--$ a short sequence of simpler operations is often faster than a single multiplication instruction.$)$