ICL Profile

Heike Jagode

Heike Jagode

Research Associate Professor

Education

  • Ph.D. in Computer Science, University of Tennessee, Knoxville
  • M.Sc. in High-Performance Computing, University of Edinburgh EPCC, Scotland, UK
  • M.Sc. in Applied Techno-Mathematics, University of Applied Sciences Mittweida, Germany
  • B.Sc. in Applied Mathematics, University of Applied Sciences Mittweida, Germany

Bio

Heike Jagode became a part of the Innovative Computing Laboratory (ICL) in 2008. Currently, she holds the position of Research Associate Professor at ICL, which is part of the Tickle College of Engineering at the University of Tennessee - Knoxville (UTK). She specializes in High Performance Computing (HPC) and the efficient utilization of advanced computer architectures. Her primary focus is on developing methods and tools for performance analysis, tuning, and energy efficiency of parallel scientific applications.

Heike’s Ph.D. research (advisor: Jack Dongarra) focused on a multi-disciplinary effort to develop Dataflow Programming Paradigms for Computational Chemistry Methods to make these methods compatible with next-generation task scheduling systems like StarPU or PaRSEC. In her research, she converted the state-of-the-art NWChem Coupled Cluster methods into different dataflow versions and demonstrated the benefits of dataflow-based task execution over coarse grain parallelism in terms of scalability, resource utilization, and programmability.

Research

  • High-Performance Computing
  • Computer Architectures
  • Performance Analysis and Tuning
  • Performance Modeling
  • Power and Energy Monitoring and Capping
  • Dataflow Programming Paradigms

Projects

  • ANACIN-X: Analysis and modeling of Non-determinism and Associated Costs IN eXtreme scale applications
  • DESC: Development of Exascale Software for Heterogeneous and Interfacial Catalysis
  • ECP: Exascale Computing Project: Exa-PAPI: The Exascale Performance Application Programming Interface
  • PAPI: The Performance API
  • PULSE: PAPI Unifying Layer for Software-defined Events

Past

  • CoDAASH: A Co-Design Approach for Advances in Software and Hardware
  • POINT: Productivity from Open, INtegrated Tools
  • SuPER: Institute for Sustained Performance, Energy, and Resilience