Of or pertaining to a floating-point number having the least precision among two or more options commonly offered by a programming language, such as single-precision versus double-precision.