Workshops Archive

Peachy Parallel Assignments (EduHPC 2026)


Workshop: EduHPC-26: Workshop on Education for High Performance Computing

Authors: Hamdan Alabsi (Bowie State University), Nathan Bean (Kansas State University), Uma Chinta (Texas Tech University), Hoda El Sayed (Bowie State University), Diego García-Álvarez and Arturo Gonzalez-Escribano (Universidad de Valladolid), Lisa L. Lacher (University of Houston Clear Lake), Rand Obeidat and Sriram Srinivasan (Bowie State University), Yuri Torres (Universidad de Valladolid), Richard Weiss (Evergreen State College), and David P. Bunde (Knox College)

Abstract: Peachy Parallel Assignments are tested and ready-to-adopt assignments to teach parallel and distributed computing concepts. They are selected competitively for being "cool and inspirational" so that students spend time working on them. After selection, they are presented at the Edu* series of workshops.

This article describes three assignments presented at the EduHPC 2026 workshop. The first is an unplugged assignment that introduces MapReduce concepts by having students count word occurrences in a popular children's book. In the second assignment, students parallelize a power grid simulation using multiple parallel programming abstractions (OpenMP for shared memory systems, MPI for distributed memory systems, and CUDA/HIP/OpenCL for GPUs). For the third assignment, they perform distributed fine tuning of a Large Language Model (LLM), a popular but extremely resource-intensive application.


Back to EduHPC-26: Workshop on Education for High Performance Computing Archive Listing

Back to Full Workshop Archive Listing