Qing Yi

Interested Areas

Compilers, programming Languages, high performance computing

Education

9/1996 -- 5/2002 : Ph.D in Computer Science, Rice University

9/1992 -- 6/1995 : M.S. in Computer Science, Institute of Computing Technology,  Chinese  Academia Cinica

9/88 -- 7/92 : B.S. in Computer Science, Shandong University, P.R.China

Publications

• A C++ infrastructure for Automatic Introduction and Translation of OpenMP Directives, Dan Quinlan, Markus Schordan and Qing Yi, Workshop on OpenMP Applications and Tools, Toronto, Ontario, Canada. June 26, 2003
• Improving Memory Hierarchy Performance Through Combined Loop Interchange and Multi-level Fusion, to appear in the special issue (LACSI 2002 symposium in Aug. 2002) of International Journal of High Performance Computing Applications.
• Transforming Complex Loop Nests For Locality, Qing Yi and Ken Kennedy, technical report TR02-386, Computer Science, Rice University. Feb, 2002.
• Advanced Optimization Strategies in the Rice dHPF compiler, John Mellor-Crummey, Vikram Adve, Bradly Broom, Daniel Chavarria-Miranda, RObert Fowler, Guohua Jin, Ken Kennedy and Qing Yi, in Concurrency and Computation: Practice and Experience, 14(8-9):741-767, July/August 2002.
• Transforming Loops To Recursion For Multi-Level Memory Hierarchies, Qing Yi, Vikram Adve, and Ken Kennedy, in proceedings of ACM SIGPLAN conference of Programming Language Design and Implementation. Vancouver, British Columbia, Canada. June, 2000
• High Performance Fortran Compilation Techniques for Parallelizing Scientific Codes, Vikram Adve, Guohua Jin, John Mellor-Crummey, Qing Yi, in Proceedings of SC98: High Performance Computing and Networking. Orlando, FL, Nov. 1998

Research Experiences

1/2003 - current: Post-doc associate in Center for Applied Scientific Computing, Lawrence Livermore National Laboratory.

Work in the ROSE project, a C++ infrastructure for user-directed optimizations.

6/2002 - 12/2002: Post-doc associate in Computer Science, Rice University

Continueing research in the area of automatic tunable loop optimations for high performance

9/1996 - 5/2002: Ph.D program in Computer Science, Rice University

Research direction: compiler optimizations for high performance computing

Dissertation contributions:

• Dependence hoisting: a new loop transformation that permits the direct fusion and interchange of arbitrarily nested loops. The transformation is highly effective and is inexpensive enough to be incorporated into commercial compilers.
• Computation slicing: a framework that systematically combines {\em dependence hoisting} with other inexpensive techniques to optimize arbitrary loop structures for better locality.
• Recursion transformation: the first compiler work that automatically transforms loop structures into recursive form to exploit locality simultaneously at multiple levels of the memory hierarchy.

9/1992 -6/1995: M.S program in Institute of Computing Technology, Chinese Academia Cinica

Research area: Chinese character encoding and automatic recognition.

Dissertation contribution: a new Chinese-handwriting automatic online-recognition system.

Academic Honor and Award

• Rice University graduate student fellowship
• Outstanding Graduate Student Honor and Award when getting M.S. degree in Institute of Computing Technology, Chinese Academia Cinica
• Outstanding student honor and Award through all four years' undergraduate study in Shandong University

Industrial Experiences

6/1995 - 7/1996 :  Software engineer in Beijing Zhenzhong Electronic Inc., P.R.China

Programming in C, independently developed the following software products:

• A BASIC compiler that produces object code for a handheld computer
• A BASIC-program debugging tool that simulates the handheld computer
• An automatic query/notification telephone system which automatically manages multiple telephone lines

Programming Skills

Proficient in C/C++, familiar with Java, Fortran, MPI, Pascal, Scheme, and various script languages