I am a Ph.D. student in Computer Science Dept. at
College of William & Mary under the supervision of Prof.
Program parallelization and optimization.
User-centric & App-aware Optimization for Mobile System.
I am in the job market for academic/research lab positions.
(If you know any openings, please forward them to me. Thanks!)
Parallelization of Inherently Sequential Programs
Parallel FSM Processing
Computations based on Finite State Machines (e.g.,
lexing, intrusion detection, pattern matching and decoding) are hard to parallelize due
to strong dependences among states. In this work, we explore the inherent relations between
speculation schemes and FSMs' properties.
We develop a probabilistic model-based approach to maximizing the speculation benefits for parallel FSM
Parallel HTML Parsing
In modern web browsers, HTML parser still runs sequentially. In this work, we target HTML5,
design and prototype both a pipelined and a data parallel HTML5 parsers by addressing various complexities in HTML documents, such as
informal grammar, auto-error correcting, embedded languages and dynamic inputs.
Program Behavior Prediction
[TACO13 / HiPEAC14]
Function Call Sequence Prediction
Function call sequence is a sequence of functions that are invoked during
program execution. Enabling function call sequence prediction
can directly benefit method prefetching and JIT compilation.
In this work, we introduce a new program representation to make this prediction
accurate and lightweight.
Mobile App Responsiveness Optimization
Responsiveness Model of Moible Apps
The responsiveness of mobile apps is an important factor to users' satisfication.
Existing work mainly use response delay reduction as the metric for responsiveness.
In this project, we find that this metric
does not faithfully reflect users' experience and could lead to wrong conclusions.
To avoid this, we provide a new metric, called SatScore.
Moible App Scheduling for Launch Responsiveness
Moible app scheduling plays a critical role on app launch responsiveness. In this work, we study the factors that
potentially affect the effectiveness of app scheduling. With a systematical exploration, we find the current design of mobile
app scheduler can be enhanced via several ways, which could substantially improve users' experience.
"Remember, the brick walls are there for a reason. The brick walls are
not there to keep us out. The brick walls are there to give us a chance
to show how badly we want someting. Because the brick walls are there to
stop the people who don't want it badly enough. They're there to stop the
other people." - Randy Pausch
OOPSLA14 accepts our paper on function call prediction
Ubicomp14 accepts our paper on mobile app launch responsiveness study
Dr. Hans Vandierendonck presented our work on parallel HTML parser at HiPEAC14
HiPEAC14 invites us to present our work on parallel HTML parser
TACO accepts our paper on parallel HTML parser
ASPLOS14 accepts our paper on parallel DFA processing
ASPLOS14 accepts our paper on JIT compilation order optimization
PPoPP13 accepts our paper on GPU memory optimizations
Office: Room 101C, McGlothlin-Street Hall
Computer Science Department
College of William & Mary
P.O. Box 8795 Williamsburg, VA 23187-8795
Email: zzhao [at] cs.wm.edu