Page 70 - profile2014.indd
P. 70
研究員
李丕榮 PeiZong Lee
Research Fellow
Ph.D., Computer Science, New York University
Tel: +886-2-2788-3799 ext. 1812 Fax: +886-2-2782-4814
Email: leepe@iis.sinica.edu.tw
http://www.iis.sinica.edu.tw/pages/leepe
● Research Fellow, Academia Sinica, Research Description
Taiwan (1998/1–present)
● Adjunct Associate Profes- My research interests are in compilers for scienti c applications, parallel algorithm design,
sor, Department of Electrical computer architectures, and the interplay among architectures, algorithms, and compilers.
Engineering, National Chung The challenge of implementing large scienti c applications for current parallel computers
Cheng University, Chiayi, Taiwan is to handle data locality among memory hierarchies, so as to avoid memory con ict on
(1993–1996)
shared memory parallel computers and to avoid communication overhead on distributed
● Associate Research Fellow, Aca- memory parallel computers. Compilers act as bridges connecting algorithms and architec-
demia Sinica, Taiwan (1989/10– tures. I am interested in studying this interdependence.
1998/1)
● Ph.D., Computer Science, New For programs that do not use indirect memory accesses (such as a subscript array of ar-
York University, United States rays or pointers) running on distributed memory parallel computers, compilers can nd
(1986/9–1989/10) data dependence relations among statements and data alignment relations among data
● M.S., Computer and decision sci- arrays. Therefore, compilers have enough information to determine data distribution, ex-
ences, National Tsing Hua Univer- ecution scheduling, and the generation of communication code using e ective communi-
sity, Taiwan (1982/9–1984/6) cation primitives. However, if a program does use a subscript array of arrays or pointers to
● B.S., Mathematics, National Tai- implement indirect memory accesses for irregular computation, compilers at the compil-
wan University, Taiwan (1975/9– ing time cannot determine either data dependence relations or data alignment relations;
1979/6) if compilers only provide naïve data distributions, generated code cannot avoid adopting
expensive communication primitives, which inevitably lengthen the execution time, and
thus degrade the performance of parallel processing.
To identify techniques for compiling irregular computation, we analyze real code for scienti c computation. This requires that we re-
search the following: unstructured mesh generation; unstructured mesh partition; Euler equation and Navier- Stokes equation solvers
for a numerical wind tunnel platform, an engine combustion platform for computing reactive ows; visualization; and the challenge of
using MPI on multi-core workstation/PC clusters to accelerate irregular computation.
Publications
1. PeiZong Lee and Zvi M. Kedem. 1988. “Synthesizing Linear- tions on Parallel and Distributed Systems, Vol. 8, No. 8, Aug.
Array Algorithms from Nested For Loop Algorithms,” in The 1997, page 825--839.
Special Issue on Parallel and Distributed Algorithms, IEEE 7. PeiZong Lee and Wen-Yao Chen. 2002. “Generating Commu-
Transactions on Computers, Vol. C-37, No. 12, Dec. 1988, nication Sets of Array Assignment Statements for Block-Cy-
page 1578--1598.
clic Distribution on Distributed Memory Parallel Computers,”
2. PeiZong Lee and Zvi M. Kedem. 1990. “Mapping Nested Parallel Computing, Vol. 28, No. 9, Sep. 2002, page 1329--
Loop Algorithms into Multidimensional Systolic Arrays,” in 1368.
IEEE Transactions on Parallel and Distributed Systems, Vol.
1, No. 1, Jan. 1990, page 64--76. 8. PeiZong Lee and Zvi M. Kedem. 2002. “Automatic Data and
Computation Decomposition on Distributed Memory Parallel
3. PeiZong Lee and Fang-Yu Huang. 1994. “Restructured Recur- Computers,” ACM Transactions on Programming Languages
sive DCT and DST Algorithms,” in IEEE Transactions on Sig- and Systems, Vol. 24, No. 1, Jan. 2002, page 1--50.
nal Processing, Vol. 42, No. 7, July 1994, page 1600--1609.
9. PeiZong Lee, Chih-Hsueh Yang, and Jeng-Renn Yang. 2004.
4. PeiZong Lee and Fang-Yu Huang. 1994. “An Efficient Prime- “Fast Algorithms for Computing Self-Avoiding Walks and
Factor Algorithm for the Discrete Cosine Transform and Its Mesh Intersections over Unstructured Meshes,” Advances in
Hardware Implementations,” IEEE Transactions on Signal Engineering Software, Vol. 35, No. 2, Feb. 2004, page 61—73.
Processing, Vol. 42, No. 8, Aug. 1994, page 1996--2005.
10. PeiZong Lee, Chien-Min Wang, and Jan-Jan Wu. 2006. “Com-
5. PeiZong Lee. 1995. “Techniques for Compiling Programs on piler and Run-time Parallelization Techniques for Scientific
Distributed Memory Multicomputers,” Parallel Computing, Computations on Distributed Memory Parallel Computers,”
Vol. 21, No. 12, Dec. 1995, page 1895--1923. included in the book High Performance Computing: Para-
digm and Infrastructure, pp. 135--181, edited by Dr. Laurence
6. PeiZong Lee. 1997. “Efficient Algorithms for Data Distribu- T. Yang and Dr. Minyi Guo, John Wiley & Sons, Inc., 2006.
tion on Distributed Memory Multicomputers,” IEEE Transac-
70 研究人員 Research Faculty