Issue No. 03 - March (2013 vol. 24)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TPDS.2012.160
Yongpeng Zhang , North Carolina State University, Raleigh
Frank Mueller , North Carolina State University, Raleigh
This paper develops and evaluates search and optimization techniques for autotuning 3D stencil (nearest neighbor) computations on GPUs. Observations indicate that parameter tuning is necessary for heterogeneous GPUs to achieve optimal performance with respect to a search space. Our proposed framework takes a most concise specification of stencil behavior from the user as a single formula, autogenerates tunable code from it, systematically searches for the best configuration and generates the code with optimal parameter configurations for different GPUs. This autotuning approach guarantees adaptive performance for different generations of GPUs while greatly enhancing programmer productivity. Experimental results show that the delivered floating point performance is very close to previous handcrafted work and outperforms other autotuned stencil codes by a large margin. Furthermore, heterogeneous GPU clusters are shown to exhibit the highest performance for dissimilar tuning parameters leveraging proportional partitioning relative to single-GPU performance.
Graphics processing unit, Arrays, Instruction sets, Kernel, Tuning, Three dimensional displays, Optimization, GPU clusters, Accelerators, GPGPU programming, stencil codes
F. Mueller and Y. Zhang, "Autogeneration and Autotuning of 3D Stencil Codes on Homogeneous and Heterogeneous GPU Clusters," in IEEE Transactions on Parallel & Distributed Systems, vol. 24, no. , pp. 417-427, 2013.