Team Phoenix-Cluster Competition Team (HPC)

2020 ~ Present

Goals

The Student Cluster Competition team will focus on putting together a proposal each Spring to compete in Supercomputing’s “Student Cluster Competition”. This competition brings togethers teams from around the world to complete a series of challenges related to high-performance computing within a set time limit and power budget. Students are expected to develop a coherent proposal and then work with team mentors and vendors to understand and deploy a working high-performance cluster. Additional goals will include installing, running, and optimizing traditional high-performance applications and benchmarks. The team will attend Supercomputing each fall to compete against other teams based on the preparation undertaken in earlier semesters in this VIP class.

Issues Involved or Addressed

All aspects of creating and using a high performance cluster computer system from designing it, deploying it, configuration, optimization, and demonstrating its use.

Methods and Technologies

  • Software development/tuning
  • Hardware architecture
  • Networking (Ethernet/Infiniband)
  • Accelerator technology (GPUs)
  • Operating System optimization
  • Parallel processing optimization
  • Computer power tuning
  • Cloud Computing
  • Scalable system management
  • Benchmarking and monitoring

Academic Majors of Interest

  • BusinessGeneral Management
  • ComputingComputational Science and Engineering
  • ComputingComputer Science
  • EngineeringComputer Engineering
  • EngineeringElectrical Engineering
  • EngineeringIndustrial Engineering
  • Other
  • SciencesApplied Physics
  • SciencesMathematics

Preferred Interests and Preparation

Required skills include: Interest in and willingness to learn basic Linux commands, experience with Makefiles, and Linux software. Experience with either Python or C programming.

Suggested skills include: Experience with basic parallel programming, Linux system maintenance, hardware and/or application profiling using gprof, nvprof, or other tools. Project Management skills are useful for team coordination and organization.

 

 

Meeting Schedule & Location

Time 
3:30-4:20
Meeting Location 
Van Leer 465
Meeting Day 
Wednesday

Team Advisors

Dr. Rich Vuduc
  • Computational Science and Engineering
Dr. Spencer Bryngelson
  • Computational Science and Engineering
Will Powell
  • Computational Science and Engineering