Supercompilers for Parallel and Vector Computers

A major challenge facing computer science is the need to develop algorithms and software tools to harness the vast power of parallel and vector computers. This book provides in-depth treatment of modern restructuring tools, called supercompilers, which detect the implicit concurrency in programs written in languages such as Fortran and transform these into equivalent concurrent code. This book also features a comprehensive discussion of the methods of program analysis and optimization, including control flow analysis, data flow analysis, and data dependence analysis. Plus, a detailed account of advanced techniques for program transformation, specifically for vectorization and parallelization in both shared-memory and distributed-memory computers.