The Floating-Point unit consists of the following functional units:
The multiply unit performs single-precision or double-precision multiplication with a 2-cycle latency and a 1-cycle repeat rate.
The divide and square-root units do single- or double-precision operations. They have long latencies and low repeat rates (20 to 40 cycles).
The add unit performs floating-point add and subtract, compare, and conversion operations. Except for Convert Integer To Single-Precision (float), all operations have a 2-cycle latency and a 1-cycle repeat rate.
Chapter Contents
Copyright 1995, MIPS Technologies, Inc. -- 29 JAN 96
Generated with CERN WebMaker