Abstract: In recent years, approximate computing has been widely adopted in error-tolerant applications. For approximate multipliers, several techniques have been proposed that truncate lower bits ...