Abstract: An algorithm for leading zero anticipation (LZA) and its implementation are vital for the performance of a high-speed floating-point adder in today's state of the art microprocessor design.