An invaluable reference for anyone designing new parallel or distributed systems. Includes detailed case studies of specific systems from Stanford, MIT, and other leading research universities. The authors emphasize performance, surveying all available techniques.
