Graduate Exam Abstract
Dylan MachovecPh.D. Preliminary
April 21, 2020, 1:00 pm - 3:00 pm
Dynamic Resource Management in Heterogeneous Environments: Maximizing Utility and Value, Energy-Constrained Scheduling, Metaheuristics, and Preemption-Capable Techniques
Abstract: The need for high performance computing (HPC) resources is rapidly expanding throughout numerous technical fields. Within a given HPC system, there are limited resources available to meet computational needs. To address this, it is important to effectively manage these resources to ensure that as much useful work as possible is completed. In one part of this research, HPC systems executing parallel jobs are considered with and without energy constraints. Additionally, HPC systems executing only serial jobs in environments where preemption is available. Dynamic resource management techniques are designed, evaluated, and compared in heterogeneous environments to assign and schedule jobs to HPC nodes. These techniques are evaluated based on system-wide performance measures (value or utility), which quantify the amount of useful work accomplished by the HPC system. Near real-time heuristics are designed to optimize performance in specific environments and in some of this research the best performing techniques are combined using intelligent metaheuristics that dynamically switch between heuristics based on the characteristics of the current environment. Resource management techniques are being developed and applied to resource management for the assignment and scheduling of heterogenous unmanned aerial vehicles (UAVs) to surveil targets, where performance is also characterized by a value-based performance measure and each UAV is constrained in its total energy consumption.
Adviser: Prof. H. J. Siegel
Non-ECE Member: Prof. Patrick Burns
Member 3: Prof. Anthony A. Maciejewski
Addional Members: Prof. Sudeep Pasricha
Bhavesh Khemka, Dylan Machovec, Christopher Blandin, Howard Jay Siegel, Salim Hariri, Ahmed Louri, Cihan Tunc, Farah Fargo, and Anthony A. Maciejewski, "Resource Management in Heterogeneous Parallel Computing Environments with Soft and Hard Deadlines," 11th Metaheuristics International Conference (MIC 2015), sponsor: CNRS (Centre National de la Recherche Scientifique), Agadir, Morocco, 10 pp., June 2015.
Dylan Machovec, Bhavesh Khemka, Sudeep Pasricha, Anthony A. Maciejewski, Howard Jay Siegel, Gregory A. Koenig, Michael Wright, Marcia Hilton, Rajendra Rambharos, and Neena Imam, "Dynamic Resource Management for Parallel Tasks in an Oversubscribed EnergyConstrained Heterogeneous Environment," 25th Heterogeneity in Computing Workshop (HCW 2016), cosponsors: IEEE Computer Society and U.S. Office of Naval Research, in the proceedings of 2016 International Parallel and Distributed Processing Symposium Workshops (IPDPSW 2016), pp. 67-78, Chicago, IL, May 2016.
Dylan Machovec, Cihan Tunc, Nirmal Kumbhare, Bhavesh Khemka, Ali Akoglu, Salim Hariri, and Howard Jay Siegel, "Value-Based Resource Management in High-Performance Computing Systems," 7th Workshop on Scientific Cloud Computing (ScienceCloud 2016), cosponsors: ACM SIGARCH (Special Interest Group on Computer Architecture) and The University of Arizona, in the proceedings of The 25th International Symposium on High Performance Parallel and Distributed Computing (HPDC `16), pp. 19-26, Kyoto, Japan, May/June 2016.
Cihan Tunc, Nirmal Kumbhare, Ali Akoglu, Salim Hariri, Dylan Machovec, and Howard Jay Siegel, "Value of Service Based Task Scheduling for Cloud Computing Systems," The IEEE International Conference on Cloud and Autonomic Computing (ICCAC'16), cosponsors: IEEE, IEEE Computing Society, and the University of Augsburg, 11 pp., Augsburg, Germany, Sep. 2016.
Dylan Machovec, Sudeep Pasricha, Anthony A. Maciejewski, Howard Jay Siegel, Gregory A. Koenig, Michael Wright, Marcia Hilton, Rajendra Rambharos, Thomas Naughton, and Neena Imam, "Preemptive Resource Management for Dynamically Arriving Tasks in an Oversubscribed Heterogeneous Computing System," 26th Heterogeneity in Computing Workshop (HCW 2017), cosponsors: IEEE Computer Society and U.S. Office of Naval Research, in the proceedings of the 2017 International Parallel and Distributed Processing Symposium Workshops (IPDPSW 2017), pp. 54-64, Orlando, FL, May 2017.
Cihan Tunc, Dylan Machovec, Nirmal Kumbhare, Ali Akoglu, Salim Hariri, Bhavesh Khemka, and Howard Jay Siegel, "Value of Service Based Resource Management for Large-Scale
Computing Systems," Cluster Computing, Vol. 20, Issue 3, pp. 2013-2030, Sep. 2017.
Nirmal Kumbhare, Cihan Tunc, Dylan Machovec, Ali Akoglu, Salim Hariri, and Howard Jay Siegel, "Value Based Scheduling for Oversubscribed Power-Constrained Homogeneous HPC
Systems," The IEEE International Conference on Cloud and Autonomic Computing (ICCAC'17), cosponsors: IEEE and IEEE Computing Society, pp. 120-130, Tucson, AZ, Sep. 2017.*
Dylan Machovec, Bhavesh Khemka, Nirmal Kumbhare, Sudeep Pasricha, Anthony A. Maciejewski, Howard Jay Siegel, Ali Akoglu, Gregory A. Koenig, Salim Hariri, Cihan Tunc,
Michael Wright, Marcia Hilton, Rajendra Rambharos, Christopher Blandin, Farah Fargo, Ahmed Louri, and Neena Imam, "Utility-Based Resource Management in an Oversubscribed EnergyConstrained Heterogeneous Environment Executig Parallel Applications," Parallel Computing, Vol. 83, pp. 48-72, Apr. 2019.
Program of Study: