Deterministic Coin Tossing With Applications to Optimal Parallel List Ranking (Classic Reprint) - Tapa blanda

Sinopsis

This book presents new deterministic parallel algorithms for list ranking a problem commonly encountered when designing parallel algorithms. The bulk of the work on the subject to date has developed deterministic algorithms based on either linear-time serial algorithms, or O(log n) time parallel algorithms using n processors. This book makes significant headway, presenting new algorithms that achieve optimal speed-up for various processor configurations. The author provides a new deterministic coin tossing technique for breaking symmetric situations in a random-like fashion. This technique is applied to the list-ranking problem to devise an O(log n time algorithm using n/(log n) processors. Additionally, this book presents algorithms that achieve optimal speed-up for all practical purposes, as well as an algorithm that uses n processors to achieve a runtime of O(log n) time, disproving a longstanding conjecture in the field. The author's insights into deterministic parallel algorithms and their applications to the list-ranking problem are significant, and this book will be of great interest to researchers and practitioners working in parallel computing and algorithm design.

"Sinopsis" puede pertenecer a otra edición de este libro.