| 摘要 |
An apparatus and method for performing high performance division by implementing a sliced design configuration wherein at least several of the most significant bits of the dividend are located in a master slice and the remaining bits of the dividend are located in at least one slave slice. Although data to be passed from the slave slices to the master slice trails the master slice's data by two cycles, the master can still determine its next operation by uniquely employing a redundant division algorithm.
|
