Task Mapping for Emerging Network Topologies

This project aims to improve the performance of large scientific simulations on next-generation high-performance computing (HPC) systems by developing new strategies for task mapping, the assignment of specific parts of an application to each of the many processing nodes in an HPC system. The specific network topology studied is Dragonfly, two variations of which are used in commercial systems. By improving application performance on next-generation HPC systems, the project will help realize the full potential of these powerful systems.

Publications and presentations

In this project, we created the following major products:

• Shortest paths in Dragonfly systems.
  Written with Ryland Curtsinger (Knox `19).
  Proceedings of the 5th IEEE International Workshop of High-Performance Interconnection Networks in the Exascale and Big-Data Era (HiPINEB), 2019.

• Brief annoucement: Coloring-based task mapping for Dragonfly systems.
  Written with Ink Chinavinijkul (Knox `18), Jacob Newcomb (Knox `20), and Lingzhi Xi (Knox `18). Proceedings of the 30th Annual ACM Symposium on Parallelism in Algorithms and Architectures (SPAA), pages 91-93, 2018.

• Improving Valiant routing for Slim Fly networks.
  Written with Deyu Han (Knox `16) and Zhaofeng Wang (Knox `17).
  Proceedings of the 10th International Workshop on Parallel Programming Models and Systems Software for High-End Computing (P2S2), 2017.

• Task mapping for emerging network topologies.
  Poster presented at the NSF CSR PI meeting, 2017.

• New link arrangements for Dragonfly networks.
  Written with Madison Belka (Knox `17), Myra Doubet (Knox `19), Sofia Meyers (Knox `17), Rosemary Momoh (Knox `17), and David Rincon-Cruz (Knox `17).
  Proceedings of the 3rd IEEE International Workshop on High-Performance Interconnection Networks in the Exascale and Big-Data Era (HiPINEB), 2017.

• Comparing global link arrangements for Dragonfly networks.
  Written with Emily Hastings (Knox `16), David Rincon-Cruz (Knox `17), Marc Spehlmann (Knox `15), Sofia Meyers (Knox `17), Anda Xu (Knox `15), and Vitus J. Leung.
  Proceedings of IEEE Cluster, pages 361-370, 2015.

All the students involved in the project have also presented their work at Knox and several of them has presented in other venues focused on undergrad research.

This material is based upon work supported by the National Science Foundation under Grant CNS-1423413.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.