A floating-point literal is a notation for representing real numbers. The assembler provides both decimal and hexadecimal floating-point notations for representing real numbers.