#2.1 Serial hardware and software #2.2 von Neumann bottleneck #2.3 MODIFICATIONS TO THE VON NEUMANN MODEL #2.4 Cache mappings #2.5 Virtual memory #2.6 Pipelining #2.7 Multiple Issue #2.8 Hardware multithreading #2.9 PARALLEL HARDWARE #2.10 Shared Memory System #2.11 PARALLEL SOFTWARE #2.12 Input and Output #2.13 PERFORMANCE #2.14 Example #2.15 Foster’s methodology

#3.1 Identifying MPI processes #3.2 Compilation #3.3 TRAPEZOIDAL RULE IN MPI #3.4 Parallelizing the Trapezoidal Rule #3.5 First version #3.6 Dealing with I/O #3.7 Input #3.8 MPI_Reduce #3.9 Collective vs. Point-to-Point Communications #3.10 MPI_Allreduce #3.11 Data distributions #3.12 Scatter #3.13 Gather #3.14 Allgather #3.15 An MPI matrix-vector multiplication function #3.16 Elapsed parallel time #3.17 MPI_Barrier #3.18 Safety in MPI programs

#4.1 Processes and Threads #4.2 Global variables #4.3 Pthreads matrix-vector multiplication #4.4 CRITICAL SECTIONS #4.5 Possible race condition #4.6 Mutexes #4.7 PRODUCER-CONSUMER SYNCHRONIZATION AND SEMAPHORES #4.8 Barrier #4.9 Condition Variables

#5.1 ميد1 #5.2 ميد1 #5.3 ميد2 #5.4 ميد2 #5.5 ميد2 #5.6 Global Sum #5.7 Cache Memory #5.8 Speedup #5.9 Speedup

#6.1 OpenMP #6.2 clause #6.3 THE TRAPEZOIDAL RULE #6.4 SCOPE OF VARIABLES #6.5 Reduction operators #6.6 Estimating #6.7 SCHEDULING LOOPS