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

Inhaltsangabe

Explore how to design and use massive parallel systems that run at scale. This work surveys architectures, networks, and software ideas for coordinating many processors to solve hard problems faster, with a focus on practical implementation and real?world constraints.

This edition examines core questions in parallel computing, from how processors connect to how programs synchronize. It explains why dataflow concepts and applicative languages are being explored, and how different network patterns support fast, scalable computation. The discussion stays grounded in what is technologically feasible while outlining models that help researchers and engineers think about the future of very large scale parallel machines.

• How interconnection networks affect speed, reliability, and hardware cost.


• Ways to coordinate many processors using simple, scalable primitives like replace-add and concurrent queues.


• Comparisons of network variants such as hypercubes, shuffle networks, and cube-connected cycles with practical tradeoffs.


• Strategies for implementing, simulating, and deploying dataflow?based programming on parallel hardware.

Ideal for readers who want a clear, accessible view of parallel computer design, from theory to hardware considerations, in a single coherent reference for very large systems.

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