Cu-T-Pi Revised: An Updated Model Supercomputer for Parallel Computing Pedagogy

  • James WolferEmail author
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 1135)


Cu-T-Pi, named for the CUDA, Nvidia TK1, and Raspberry Pi technology included, is a heterogeneous model supercomputer. Used as a pedagogic tool for teaching high-performance parallel computing, this model supports the major programming paradigms used in modern supercomputing. This work describes a complete remake of the original computer as a hardware and performance refresh, along with augmentation to support embedded Deep Learning.


High-Performance Computing Parallel computing Hardware model Deep-learning inference Pedagogy 



The author gratefully acknowledges the Center for Parallel and Distributed Computing for their kind grant supporting Parallel and Distributed Computing curriculum development.


  1. 1.
    ACM/IEEE-CS Joint Taskforce: Computer Science Curricula 2013 Final Report 0.9, Pre-release version, October 2013.
  2. 2.
    Prasad, S. et al.: NSF/IEEE-TCPP Curriculum Initative on Parallel and Distributed Computing – Core Topics for Undergraduates (2012).
  3. 3.
    Wolfer, J.: A heterogeneous supercomputer model for high-performance parallel computing pedagogy. In: Proceedings of IEEE EDUCON (2015)Google Scholar
  4. 4.
    Wolfer, J.: A model supercomputer for instructional support. In: Experiment@International Conference Demonstration Session (2016)Google Scholar
  5. 5.
    Adams, J., Caswell, J., Matthews, S., Peck, C., Shoop, E., Toth, D., Wolfer, J.: The microcluster showcase: 7 inexpensive beowulf clusters for teaching PDC. In: ACM SIGCSE (2016)Google Scholar
  6. 6.
    Adams, J., Matthews, S., Shoop, E., Toth, D., Wolfer, J.: Using inexpensive microclusters and accessible materials for cost-effective parallel and distributed computing education. J. Comput. Sci. Educ. 8(3), 2–10 (2017)CrossRefGoogle Scholar
  7. 7.
  8. 8.
  9. 9.
  10. 10.
    Raspberry Pi Foundation.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
    Top500: Top 500 Supercomputer Sites.
  16. 16.
    Heroux, M.A., Dongarra, J.: Toward a new metric for ranking high performance computing systems, SAND2013-4744, 10 June 2013.
  17. 17.
    Movidius: Getting Started Guide for the Movidius Neural Compute Stick, Movidius an Intel Company, July 2017Google Scholar
  18. 18.
    Keeler, W., Wolfer, J.: A Raspberry Pi cluster and Geiger counter supporting random number acquisition in the cs operating systems class. In: Proceedings of the International Conference on Remote Engineering and Virtual Instrumentation, February 2015Google Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Indiana University South BendSouth BendUSA

Personalised recommendations