Networks and Algorithms for Very Large Scale Parallel Computation (Classic Reprint) - Softcover

Inhaltsangabe

Explore how thousands of processors can work in concert. Discover practical ideas for architectures, networks, and software that coordinate massive parallelism.

This book surveys the design of highly parallel computers and the software needed to run them. It introduces concepts like paracomputers, dataflow execution, and synchronization techniques that help many processors share work and memory efficiently.

• Foundations of interconnection networks and asymptotically optimal patterns for large-scale computation
• A practical look at replace-add synchronization and its impact on parallel algorithms
• Hardware and software approaches to concurrent queues, scheduling, and dataflow execution
• Applications and simulations that illustrate scalable performance in real-world codes

Die Inhaltsangabe kann sich auf eine andere Ausgabe dieses Titels beziehen.