Parallel Algorithms

Academic (University) preparation with worked examples, written practice and independent learning assessments.

Independent university-level study of Parallel Algorithms. This original curriculum draws on public academic subject descriptions; it is not an official faculty syllabus, university credit, accreditation or an Ivy League qualification. Topics covered: Work and span; Parallel models; Speedup bounds; Task decomposition. Parallel scan; Reduction; Sorting; Selection. Graph traversal; Connected components; Parallel shortest paths; Graph partitioning. Matrix multiplication; Sparse kernels; Parallel factorization; Communication costs. Shared-memory algorithms; Work stealing; Contention; False sharing. GPU execution; Memory coalescing; Warp divergence; Kernel decomposition. Distributed collectives; Load balancing; Communication avoidance; Scalability analysis. Performance models; Strong scaling; Weak scaling; Parallel algorithm project. Projects use safe simulations, public datasets or textual analysis; specialist subject review is required. Includes lessons, supplementary YouTube tutorials, practical exercises, quizzes and free/paid practice tests. Erudex is independent, not affiliated with the source institution or awarding body. Its 80% learning checks are not official grades or university credit. Content requires qualified subject review before use as a complete teaching programme.