High performance computing: programming and applications

Multicore ArchitecturesMEMORY ARCHITECTURESSE INSTRUCTIONS HARDWARE DESCRIBED IN THIS BOOKThe MPP: A Combination of Hardware and Software TOPOLOGY OF THE INTERCONNECT INTERCONNECT CHARACTERISTICS THE NETWORK INTERFACE COMPUTERMEMORY MANAGEMENT FOR MESSAGESHOW MULTICORES IMPACT THE PERFORMANCE OF THE INTERCONNECTHow Compilers Optimize Programs MEMORY ALLOCATION MEMORY ALIGNMENT VECTORIZATIONPREFETCHING OPERANDS LOOP UNROLLING INTERPROCEDURAL ANALYSISCOMPILER SWITCHES FORTRAN 2003 AND ITS INEFFICIENCIES SCALAR OPTIMIZATIONS PERFORMED BY THE COMPILERParallel Programming Paradigms HOW CORES COMMUNICATE WITH EACH OTHERMESSAGE PASSING INTERFACEUSING OPENMPPOSIX THREADS PARTITIONED GLOBAL ADDRESS SPACE LANGUAGES (PGAS)COMPILERS FOR PGAS LANGUAGES THE ROLE OF THE INTERCONNECTA Strategy for Porting an Application to a Large MPP System GATHERING STATISTICS FOR A LARGE PARALLEL PROGRAMSingle Core Optimization MEMORY ACCESSING VECTORIZATIONSUMMARYParallelism across the Nodes APPLICATIONS INVESTIGATEDLESLIE3D PARALLEL OCEAN MODEL (POP) SWIM S3D LOAD IMBALANCE COMMUNICATION BOTTLENECKSOPTIMIZATION OF INPUT AND OUTPUT (I/O)Node Performance APPLICATIONS INVESTIGATED WUPWISE SWIM MGRID APPLU GALGEL APSI EQUAKE FMA-3D ARTAMMPSUMMARYAccelerators and Conclusion ACCELERATORSCONCLUSIONAppendix A: Common Compiler DirectivesAppendix B: Sample MPI Environment VariablesReferencesIndexExercises appear at the end of each chapter.