Publisher's Synopsis
Excerpt from Simd Algorithms for D Arrays in Shuffle Networks
Our algorithms correspond to a simd Perfect Shuffle connected machine (ps) For simplicity, although it is not crucial, we assume that P is a power of two. The processors are ordered left to right and are numbered 0 P - l accordingly, so we denote the j 'th processor pej. Subgroups of processors are referred to in a natural manner: left half, right half, odd, even etc. There are three kinds of communication steps: shuffle unshuffle and exchange (ex) and a processor internal (fixed-size) computation. All of these operations take one time step.
In any parallel network machine we have two possible working modes One in which each processor computes a local result of a sequential computation using its local data. Then an inter-processor global algorithm is carried out to compute the final result. In other words, this mode involves separate sequential work by pe's followed by a global parallel step. It is associated with column order of the array: first each column is processed, only then a row of partial results is being processed.
About the Publisher
Forgotten Books publishes hundreds of thousands of rare and classic books. Find more at www.forgottenbooks.com
This book is a reproduction of an important historical work. Forgotten Books uses state-of-the-art technology to digitally reconstruct the work, preserving the original format whilst repairing imperfections present in the aged copy. In rare cases, an imperfection in the original, such as a blemish or missing page, may be replicated in our edition. We do, however, repair the vast majority of imperfections successfully; any imperfections that remain are intentionally left to preserve the state of such historical works.
