Warp-level divergence in GPUs: Characterization, impact, and mitigation

Proceedings - International Symposium on High-Performance Computer Architecture

Volumn , Issue , 2014, Pages 284-295

  Xiang, Ping ^a   Yang, Yi ^b   Zhou, Huiyang ^a  

^a North Carolina State University   (United States)

^b NEC LABORATORIES AMERICA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

COMPUTER ARCHITECTURE; COMPUTER GRAPHICS; HARDWARE; NATURAL RESOURCES MANAGEMENT; PROGRAM PROCESSORS; RESOURCE ALLOCATION; SUPERCOMPUTERS;

ARCHITECTURAL SUPPORT; CRITICAL RESOURCES; GRAPHICS PROCESSING UNITS; HARDWARE OVERHEADS; RESOURCE MANAGEMENT; RESOURCE MANAGEMENT SCHEMES; STREAMING MULTIPROCESSORS; THREAD-LEVEL PARALLELISM;

WEAVING;

EID: 84903999614     PISSN: 15300897     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1109/HPCA.2014.6835939     Document Type: Conference Paper

References (26)

1. A. Bakhoda, et al., "Analyzing CUDA Workloads Using a Detailed GPU Simulator, " ISPASS-2009, 2009
2. N. Brunie, et al., "Simultaneous Branch and Warp Inter-weaving for Sustained GPU Performance"ISCA-39, 2012.
3. CUDA programming guide
4. J. D. Collins, et al., "Control flow optimization via dynamic reconvergence prediction, " MICRO-37, 2004
5. S. Che, et al., "Rodinia: A Benchmark Suite for Hetero-geneous Computing, " IISWC-2009, 2009.
6. G. Diamos, et al., "SIMD Re-Convergence at Thread Frontiers, " MICRO-44, 2011.
7. W. W. Fung, et al., "Thread Block Compaction for Efficient SIMT Control Flow, " HPCA-17, 2011.
8. W. Fung, et al., "Dynamic Warp Formation and Schedul-ing for Efficient GPU Control Flow, " MICRO-40, 2007.
9. M. Gebhart, et al., "Energy-efficient mechanisms for managing thread context in throughput processors, " ISCA-38, 2011
10. T. D. Han, et al., "Reducing Branch Divergence in GPU Programs, " GPGPU-4, 2011.
11. IMPACT Research Group. The Parboil Benchmark Suite.
12. J. Leng, et al., "GPUWattch: Enabling Energy Optimizations in GPGPUs, " ISCA-40, 2013.
13. D. Kirk, et al., "Programming Massively Parallel Processors: A hand-on Approach" Morgan Kaufmann, 2012.
14. J. Lee, et al., "TLP-Aware Cache Management Schemes for a CPU-GPU Heterogeneous Architecture", HPCA-18, 2012.
15. A. Mohammad, et al., "Warped Register File: A Power Efficient Register File for GPGPUs, " HPCA-19, 2013.
16. J. Meng, et al., "Dynamic Warp Subdivision for Integrated Branch and Memory Divergence Tolerance"ISCA-37, 2010.
17. V. Narasiman, et al., " Improving GPU Performance via Large Warps and Two-Level Warp Scheduling, " MICRO-44, 2011.
18. NVIDIA. Fermi: NVIDIA's Next Generation CUDA Compute Architecture, 2011.
19. NVIDIA. CUDA C/C++ SDK Code Samples, 2011. http://developer.nvidia.com/ gpu-computing-sdk, 2011
20. Minsoo Rhu, et al., "The Dual-Path Execution Model for Efficient GPU Control Flow, " HPCA-19, 2013.
21. Minsoo Rhu, et al., "CAPRI: Prediction of Compaction-Adequacy for Handling Control-Divergence in GPGPU Architectures, " ISCA-39, 2012.
22. T. Rogers, et al., "Cache-Conscious Wavefront Scheduling, " MICRO-45, 2012.
23. D. Tarjan, et al., "On demand register allocation and deallocation for a multithreaded processor." US Patent 20110161616 A1. 2009
24. J. Tan, et al., "RISE: Improving Streaming Processors Reliability against Soft Errors in GPGPUs" PACT-21, 2012
25. J. Tan, et al., "Analyzing Soft-Error Vulnerability on GPGPU Microarchitecture, " IISWC-2011, 2011.
26. Y. Yang, et al., "Shared Memory Multiplexing: A Novel Way to Improve GPGPU Throughput, " PACT-21, 2012.