(+61 2) 9385 4941
pengd at cse dot unsw dot edu dot au
I am a postdoctoral researcher working with Professor Jingling Xue at School of Computer Science and Engineering, The University of New South Wales.
2009 - 2013 University of New South Wales (Ph.D.)
2005 - 2009 University of Science and Technology Beijing (Combined M.Sc.-Ph.D.)
2000 - 2004 Xian Jiaotong University (B.E.)
Program Analysis: Flow-Sensitivity, Context-Sensitivity, Field-Sensitivity. More efficient, scalable and precise alias analysis algorithms for bug detection on multi-threaded practical problems.
Compilation Techniques for Heterogeneous Systems: High-level programming language design and implementation for heterogeneous systems.
Compiler optimizations for lower-lever native heterogeneous languages such as CUDA, OpenCL. Automatic parallelization and performance optimization for heterogeneous systems.
Parallel Programming & Optimisation: Fine-grained parallel algorithm for iterative algorithms, GPGPU and CUDA optimization, MPI process optimizing mapping for hierarchical multi-core clusters.
Peng Di, Yulei Sui and Jingling Xue. Sparse Flow-Sensitive Pointer Analysis for Multithreaded C Programs.
In the 2016 International Symposium on Code Generation and Optimization (CGO'16), Barcelona, Spain, 2016.Artifacts
Peng Di, Yulei Sui, Ding Ye and Jingling Xue. Region-Based May-Happen-in-Parallel Analysis for C Programs.
In the 2015 International Conference on Parallel Processing (ICPP'15), Beijing, China, 2015.
Peng Di, Ding Ye, Yu Su, Yulei Sui and Jingling Xue. Automatic Parallelization of Tiled Loop Nests with Enhanced Fine-Grained Parallelism on GPUs.
In the 2012 International Conference on Parallel Processing (ICPP'12), Pittsburgh, USA, 2012.
Jilin Zhang, Liting Zhu, Jie Mao, Jian Wan and Peng Di. An Efficient Parallel Implementation for Three-dimensional Incompressible Pipe Flow based on SIMPLE.
In the 12th IEEE/ACM International Symposium on Cluster, Cloud, and Grid Computing (CCGrid'12), pages 660-667, Ottawa, Canada, 2012.
Peng Di and Jingling Xue. Model-Driven Tile Size Selection for DOACROSS Loops on GPUs.
In the 17th International European Conference on Parallel and Distributed Computing (Euro-Par'11), pages 401-412, Bordeaux, France, 2011.
Peng Di, Qing Wan, Xuemeng Zhang, Hui Wu and Jingling Xue. Toward Harnessing DOACROSS Parallelism for Multi-GPGPUs.
In the 2010 International Conference on Parallel Processing (ICPP'10), pages 40-50, San Diego, USA, 2010.
Peng Di, Jingling Xue, Changjun Hu and Jingjing Zhou. A Cache-Effcient Parallel Gauss-Seidel Solver with Alternating Tiling.
In the 15th International Conference on Parallel and Distributed Systems (ICPADS'09), pages 244-251, Shenzhen, China, 2009.
Jilin Zhang, Junfeng Yuan, Jian Wan, Peng Di. Efficient Parallel Implementation of Incompressible Pipe
Flow Algorithm Based on SIMPLE.
Concurrency and Computation: Practice and Experience, 2013.
Peng Di, Hui Wu, Jingling Xue, Feng Wang and Canqun Yang. Parallelizing SOR for GPGPUs Using Alternate Loop Tiling.
Parallel Computing, volume 38, issues 6-7, pages 310-328, June-July 2012.
Peng Di, Automatic Parallelization of Tiled Stencil Loop Nests on GPUs.
Ph.D. Thesis, 2013.
Program Committee for:
HPDIC 2012 - 2015 (in conjunction with IEEE IPDPS.)
FM-S&C 2011 (in
conjunction with IEEE ICWS 2011, SCC2011, CLOUD 2011, and SERVICES 2011.)
Conference Reviewer for:
Journal Reviewer for: