NCSI: Regional University Learners Programming Cohort

Status	Completed
Mentor Name	Jeremy Evert
Mentor's XSEDE Affiliation	Campus Champion
Mentor Has Been in XSEDE Community	4-5 years
Project Title	Regional University Learners Programming Cohort
Summary	Faculty at a regional university will mentor Learners for the XSEDE EMPOWER Learners program to enable them to compete in the state HPC competition. Through weekly meetings, faculty will guide the students through multiple modules from the SHODOR site and work with multiple HPC systems. Faculty will also coordinate meetings with other regional and national leaders in HPC education.
Job Description	Student duties include learning programming in a supercomputing environment for the purpose of scientific programming. They will examine multiple algorithms while working through several of the Petascale materials as learning exercises. They will experience multiple architectures by running parallel code from simple systems, such as virtual cluster computers built on their laptop, up to larger systems available, such as Bridges, Frontera, and Stampede 2.
Computational Resources	Student duties include learning programming in a supercomputing environment for the purpose of scientific programming. They will examine multiple algorithms while working through several of the Petascale materials as learning exercises. They will experience multiple architectures by running parallel code from simple systems, such as virtual cluster computers built on their laptop, up to larger systems available, such as Bridges, Frontera, and Stampede 2.
Contribution to Community
Position Type	Learner
Training Plan	The basic training plan is to use the state HPC competition as a basis for learning HPC. Students will begin by understanding the problem to be solved, working through basic examples by hand, and then build up to serial solutions. Students will measure time to complete the tasks serially, and then examine parallelization and other libraries. Students will compete in the state HPC competition, and be encouraged to present their learning at the state Supercomputing Symposium. Learning Modules for students to work through: Learning Module: A Beginner's Guide to High-Performance Computing Learning Module: Parallelization: Area Under a Curve "Learning Module: Introduction to OpenMP " Learning Module: GalaxSee HPC Module 1: The N-Body Problem, Serial and Parallel Simulation Learning Module: Multithreading and Multiprocessing Learning Module: Techniques and Technologies Learning Module: An Introduction to Scientific Computing Systems for students to work with: Raspberry Pi Cluster Laptop based Cluster TACC Fronteria Programming Exercises: Programming: Sequential: Write a sequential guess and check solution for the traveling salesman problem Programming: Parallel: Write a parallel hello world program in MPI " Programming: Sequential: Write a sequential solution for the determinant Programming: Sequential: Write an MPI version of the guess and check solution for TSP" Programming: Parallel: Wrie an MPI version with openMP for the determinant Programming: Parallel: Write an MPI TSP code Programming: Parallel: Write an MPI with openMP code for TSP " Programming: Parallel: Test a library solution for the determinant on one node." " Programming: Parrallel: Test a library solution for the determinant on a full raspberry pi cluster"
Student Prerequisites/Conditions/Qualifications	Every student that works with me will be required to: Attend the weekly state wide phone call with cyber infrastructure professionals every week they work on the project. Arrange a 1 hour office hour time with me to go over their progress and discuss barriers they are having Arrange a 1 hour class time with me and the rest of the cohort to come together as a group and talk through the weekly assignment. This meeting time might also be used to host virtual guest speakers, such as center directors, or one of their team members. Turn in a final 5-7 page report that details their experience, what they enjoyed the most, learned the most from, and which parts they wish they had more time with. They should turn in their best time to solve a 5,000 by 5,000 determinant on a Raspberry Pi cluster. They should turn in their best distance a standard city traveling salesman problem of their choice from TSPLIB. Turn in a 3-5 page research proposal for the topic they would like to go research next. This should include a list of papers on this subject, some key researchers, and which supercomputers are running code for this area.
Duration	Semester
Start Date	01/14/2020
End Date	04/07/2020