Shilei Tian

Work Experience

• Software Engineer, AMD, MA, 2024/01 – Present
• Research Intern, Microsoft Research, NY (remote), 2023/05 – 2023/08
• Compiler Engineer Intern, Cray Compiler Team, Hewlett Packard Enterprise, NY (remote), 2022/05 – 2022/08
• Software Engineer, Intel Compilers and Languages (ICL), Intel, Shanghai, 2018/04 – 2019/08
• Software Engineer Intern, Machine Learning and Translation (MLT), Intel, Shanghai, 2017/03 – 2018/04

Education

• Ph.D., Computer Science, Stony Brook University (SBU), 2019/08 – 2023/12
  • Thesis: Efficient Development and Execution of OpenMP on GPUs
• M.Eng., Computer Technology, Shanghai Jiao Tong University (SJTU), 2015/09 – 2018/03
• B.Eng., Computer Science and Technology, Xi'an University of Posts and Telecommunications (XUPT), 2010/09 – 2014/07

Publications

Please note that the copyrights of the following papers belong to the corresponding publishers.

• Renyuan Liu, Yuyang Leng, Shilei Tian, Shaohan Hu, Chun-Fu (Richard) Chen, Shuochao Yao, “DynaSpa: Exploiting Spatial Sparsity for Efficient Dynamic DNN Inference on Devices”, in ACM Conference on Embedded Networked Sensor Systems (SenSys), November 4th – 7th, 2024, Hangzhou, China (Honorable Mention Award) [PDF]
• Rahulkumar Gayatri, Shilei Tian, Stephen Olivier, Johannes Doerfert, Eric Wright, “Leveraging LLVM OpenMP GPU Offload Optimizations for Kokkos Applications”, in IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC), December 18th – 21st, 2024, Bengaluru, India (Distinguished Paper Award) [PDF]
• Smeet Chheda, Gaurav Verma, Shilei Tian, Barbara Chapman, Johannes Doerfert, “Evaluating Tuning Opportunities of the LLVM/OpenMP Runtime”, in Workshop on the LLVM Compiler Infrastructure in HPC (LLVM-HPC), November 18th, 2024, Atlanta, Georgia, USA [PDF]
• Shilei Tian, Tom Scogland, Barbara Chapman, Johannes Doerfert, “OpenMP Kernel Language Extensions for Performance Portable GPU Codes”, in Workshop on the LLVM Compiler Infrastructure in HPC (LLVM-HPC), November 12th, 2023, Denver, Colorado, USA [PDF]
• Shilei Tian, Barbara Chapman, Johannes Doerfert, “Exploring the Limits of Generic Code Execution on GPUs via Direct (OpenMP) Offload”, in International Workshop on OpenMP (IWOMP), September 13th – 15th, 2023, Bristol, UK [PDF]
• Shilei Tian, Barbara Chapman, Johannes Doerfert, “Maximizing Parallelism and GPU Utilization For Direct GPU Compilation Through Ensemble Execution”, in Workshop on LLVM in Parallel Processing (LLPP), August 7th, 2023, Salt Lake City, Utah, USA [PDF]
• Eric Wright, Johannes Doerfert, Shilei Tian, Barbara Chapman, Sunita Chandrasekaran, “Implementing OpenMP's SIMD Directive in LLVM's GPU Runtime”, in International Conference on Parallel (ICPP), August 7th – 10th, 2023, Salt Lake City, Utah, USA [PDF]
• Wenbin Lu, Shilei Tian, Tony Curtis, and Barbara Chapman, “Extending OpenMP and OpenSHMEM for Efficient Heterogeneous Computing”, in Parallel Applications Workshop, Alternatives To MPI+X (PAW-ATM), November 14th, 2022, Dallas, Texas, USA [PDF]
• Shilei Tian, Joseph Huber, Konstantinos Parasyris, Barbara Chapman, Johannes Doerfert, “Direct GPU Compilation and Execution for Host Applications with OpenMP Parallelism”, in Workshop on the LLVM Compiler Infrastructure in HPC (LLVM-HPC), November 13th, 2022, Dallas, Texas, USA [PDF]
• Shilei Tian, Joseph Huber, John Tramm, Barbara Chapman, Johannes Doerfert, “Just-in-Time Compilation and Link Time Optimization for OpenMP Target Offloading”, in International Workshop on OpenMP (IWOMP), September 27th – 30th, 2022, Chattanooga, Tennessee, USA [PDF]
• Johannes Doerfert, Atmn Patel, Joseph Huber, Shilei Tian, Jose M Monsalve Diaz, Barbara Chapman, Giorgis Georgakoudis, “Co-Designing an OpenMP GPU Runtime and Optimizations for Near-Zero Overhead Execution”, in IEEE International Parallel & Distributed Processing Symposium (IPDPS), May 30th – June 3rd, 2022, Lyon, France [PDF]
• Joseph Huber, Melanie Cornelius, Giorgis Georgakoudis, Shilei Tian, Jose M Monsalve Diaz, Kuter Dinel, Barbara Chapman, Johannes Doerfert, “Efficient Execution of OpenMP on GPUs”, in International Symposium on Code Generation and Optimization (CGO), April 2nd – 6th, 2022, Seoul, South Korea [PDF]
• Seonmyeong Bak, Colleen Bertoni, Swen Boehm, Reuben Budiardja, Barbara Chapman, Johannes Doerfert, Markus Eisenbach, Hal Finkel, Oscar Hernandez, Joseph Huber, Shintaro Iwasaki, Vivek Kale, Paul R.C. Kent, JaeHyuk Kwack, Meifeng Lin, Piotr Luszczek, Ye Luo, Buu Pham, Swaroop Pophale, Kiran Ravikumar, Vivek Sarkar, Thomas Scogland, Shilei Tian, P.K. Yeung, “OpenMP Application Experiences: Porting to Accelerated Nodes”, in Parallel Computing, October 23rd, 2021 [PDF]
• Atmn Patel, Shilei Tian, Johannes Doerfert, Barbara Chapman, “A Virtual GPU as Developer-Friendly OpenMP Offload Target”, in Workshop on LLVM in Parallel Processing (LLPP), August 9th, 2021, Chicago, IL, USA [PDF]
• Shilei Tian, Jon Chesterfield, Johannes Doerfert, Barbara Chapman, “Experience Report: Writing A Portable GPU Runtime with OpenMP 5.1”, in International Workshop on OpenMP (IWOMP), September 15th – 17th, 2021, Bristol, UK [PDF]
• Shilei Tian, Johannes Doerfert, Barbara Chapman, “Concurrent Execution of Deferred OpenMP Target Tasks with Hidden Helper Threads”, in Workshop on Languages and Compilers for Parallel Computing (LCPC), October 14th – 16th, 2020, Stony Brook, New York, USA [PDF]
• Shilei Tian, Haotian Wang, Sha Li, Fan Wu, Guihai Chen, “Trajectory-Based Multi-Hop Relay Deployment in Wireless Networks”, in International Conference on Combinatorial Optimization and Applications (COCOA), December 16th – 18th, 2017, Shanghai, China [PDF]
• Haotian Wang, Shilei Tian, Xiaofeng Gao, Lidong Wu, Guihai Chen, “Approximation Designs for Cooperative Relay Deployment in Wireless Networks”, in International Conference on Distributed Computing Systems (ICDCS), June 5th – 8th, 2017, Atlanta, Georgia, USA [PDF]

Talks

• Shilei Tian, Johannes Doerfert, “OpenMP Kernel Language Extensions for Performance Portable GPU Codes”, in LLVM Developers' Meeting, October 11th – 12th, 2023 [YouTube]
• Shilei Tian, Joseph Huber, “Direct GPU Compilation & Execution for Host Applications with OpenMP Parallelism”, in LLVM Developers' Meeting, November 8th – 9th, 2022 [YouTube]
• Shilei Tian, Joseph Huber, Johannes Doerfert, “LTO and JIT Support in LLVM OpenMP Target Offloading”, in LLVM Developers' Meeting, November 16th – 19th, 2021 [YouTube]
• Johannes Doerfert, Joseph Huber, Stefan Stipanovic, Giorgis Georgakoudis, Hamilton Tobon Mosquera, Shilei Tian, “(OpenMP) Parallelism Aware Optimizations”, in LLVM Developers' Meeting, October 6th – 8th, 2020 [YouTube]
• Shilei Tian, Johannes Doerfert, Barbara Chapman, “Asynchronous OpenMP Offloading on NVIDIA GPUs”, in LLVM Performance Workshop, February 23rd, 2020, San Diego, USA [Slides]

Services

• Proceeding chair, International Conference on Parallel Processing (ICPP) 2025
• Web chair, International Symposium on Code Generation and Optimization (CGO) 2025
• Co-organizer, LLVM Performance Workshop 2022, 2023, 2024, 2025
• Mentor, Google Summer of Code 2024, 2025
• Reviewer, IEEE Transactions on Parallel and Distributed Systems (TPDS) 2023, 2024
• Co-organizer, Workshop on LLVM in Parallel Processing (LLPP) 2024
• Program committee, International Workshop on LLVM for GPUs (LLVM-GPU) 2024
• Primary representative for Stony Brook University, OpenMP Architecture Review Board (ARB), 2021/05 – 2023/12
• Program committee, The International Workshop on Languages and Compilers for Parallel Computing (LCPC) 2023
• Reviewer, International Workshop on OpenMP (IWOMP) 2023
• Program committee, Workshop on LLVM in Parallel Processing (LLPP) 2023
• Artifact evaluation committee, International Symposium on Code Generation and Optimization (CGO) 2022
• Reviewer, ACM Transactions on Architecture and Code Optimization (TACO) 2021
• Web chair, International Conference on Combinatorial Optimization and Applications (COCOA) 2017