Department of Physics

18210  Development and Evaluation of Parallel Applications   

Course Webpage: https://eclass.uoa.gr/courses/D36/ 

Course Content

• Introduction to parallel computing (need and challenges).
• Parallel hardware (SIMD and MIMD systems, GPUs, interconnection networks, cache coherence and consistency, shared and distributed memory systems).
• Parallel software (processes, multitasking, threads, synchronization and communication in shared and distributed memory systems).
• Speedup, performance, cost, and scalability metrics.
• Development of parallel programs (parallel application design and evaluation methodology). Data parallelism and task parallelism.
• Programming shared memory systems with Pthreads and OpenMP, programming distributed memory systems with MPI, programming graphics processing units (CUDA), and their hybrid combination.