Operating System Considerations for Large-Scale MIMD Machines (Classic Reprint)

Jan Edler, Allan Gottlieb

ISBN 10: 1332874789 / ISBN 13: 9781332874781

Published by Forgotten Books, 2018

About this Item

Print on Demand. This book explores the design and implementation of highly parallel operating systems, pursuing the concept of critical-section-free algorithms that enable efficient coordination of tasks and resources. Providing both theoretical and practical insights, the author presents a comprehensive study of parallel processing concepts, discussing synchronization primitives, coordination algorithms, memory management techniques, scheduling policies, and more. Through a detailed examination of the NYU Ultracomputer, a prototype parallel computer, the book illustrates the challenges and solutions in building a massively parallel architecture. The author's work contributes to the understanding of parallel computer systems, offering valuable guidance for researchers and practitioners in the field. This book is a reproduction of an important historical work, digitally reconstructed using state-of-the-art technology to preserve the original format. In rare cases, an imperfection in the original, such as a blemish or missing page, may be replicated in the book.

Seller Inventory # 9781332874781_0

Bibliographic Details

• Title: Operating System Considerations for Large-Scale MIMD Machines (Classic Reprint)
• Publisher: Forgotten Books
• Publication Date: 2018
• Language: English
• ISBN 10: 1332874789
• ISBN 13: 9781332874781
• Binding: Paperback
• Condition: New
• Book Type: print-on-demand item
• Seller catalogs:
  • Audits and Surveys

About this title

Synopsis

Design principles for scalable massively parallel machines and their software

This book explains how to build hardware and software that work together to run many processors without bottlenecks, using the Ultracomputer as a guiding example. The text outlines goals like scaling performance, general purpose usefulness, and easy programmability. It shows how an operating system can support high-level parallel programming, dynamic scheduling, and parallel data management. You’ll find practical ideas for synchronization, memory management, and task coordination that aim to keep many processors busy without serial bottlenecks.

• How fetch-and-add and other coordination primitives support parallel software
• Techniques for parallel memory management and stack handling
• Strategies for scheduling, task spawning, and load balancing in a shared-memory, multi-processor setting
• Approaches to I/O, file systems, and event-based synchronization in a parallel environment

Ideal for readers interested in the hardware-software tradeoffs of parallel systems and in designing operating systems that scale with the number of processors.