Home > Events > PhD Defense - Li Tang

PhD Defense - Li Tang

Start: 4/3/2017 at 12:00PM
End: 4/3/2017 at 3:00PM
Location: 117 J Cushing Hall
Attendees: Faculty and students are welcome to attend the presentation portion of the defense.
Add to calendar:
iCal vCal

Li Tang

Dissertation Defense

April 3, 2017             12:00 pm           117 J Cushing

Adviser:  Dr. Sharon Hu

Committee:

Dr. Jeanne Cook          Dr. Peter Kogge          Dr. Michael Niemier

 

Title:
Performance and Energy Aware Workload Partitioning on Heterogeneous Platforms


Heterogeneous platforms which employ a mix of CPUs and GPUs have been widely used in many different areas such as embedded computing and high-performance computing. Such heterogeneous platforms have potential to offer higher performance at lower energy cost than homogeneous platforms. However, it is rather challenging to actually achieve the high performance and energy efficiency promised by heterogeneous platforms. One reason is that simple workload distribution of application usually cannot efficiently utilize the distinct processors of heterogeneous platforms. Another reason is that a heterogeneous platform presents a large design space for workload partitioning between different processors. To help application developers partition workload on heterogeneous platforms for exploiting performance and energy potential, we make four contributions in this dissertation. First, we study different strategies of partitioning the workload of the data assembly (DA)stage in finite element method applications. By running different workload partitions (WPs) of DA on a broad range of heterogeneous platforms, we examine the performance and energy impacts of workload partitioning for heterogeneous platforms. Second, we develop a performance model, PerDome, to estimate the performance potential of different WPs on heterogeneous platforms. Third, we build a framework, PeaPaw, to assist application developers to find a WP that has high potential leading to high performance or energy efficiency before actual implementation. The PeaPaw framework includes both analytical performance/energy models and two sets of workload partitioning guidelines. Based on the design goal (i.e., performance or energy), application developers can obtain a workload partitioning guideline and use PeaPaw to estimate the performance or energy of designed WPs on a given heterogeneous platform. Last, we enhance the ability of PeaPaw for handling more complicated application memory behaviors and integrated heterogeneous platforms in terms of performance. Our results show that PaPaw+ can provide higher accuracy of WP performance estimation and more detailed workload partitioning guidelines on both discrete and integrated heterogeneous platforms than PeaPaw.