Abstract
Several highly optimized implementations of Finite Difference schemes are discussed. The combination of vectorization and an interleaved data layout, spatial and temporal loop tiling algorithms, loop unrolling, and parameter tuning lead to efficient computational kernels in one to three spatial dimensions, truncation errors of order two to twelve, and isotropic and compact anisotropic stencils. The kernels are implemented on and tuned for several processor architectures like recent Intel Sandy Bridge, Ivy Bridge and AMD Bulldozer CPU cores, all with AVX vector instructions as well as Nvidia Kepler and Fermi and AMD Southern and Northern Islands GPU architectures, as well as some older architectures for comparison. The kernels are either based on a cache aware spatial loop or on time-slicing to compute several time steps at once. Furthermore, vector components can either be independent, grouped in short vectors of SSE, AVX or GPU warp size or in larger virtual vectors with explicit synchronization. The optimal choice of the algorithm and its parameters depend both on the Finite Difference stencil and on the processor architecture.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Datta, K., Kamil, S., Williams, S., Oliker, L., Shalf, J., Yelick, K.: Optimization and performance modeling of stencil computations on modern microprocessors. SIAM Rev. 51(1), 129–159 (2009)
Micikevicius, P.: 3D finite difference computation on GPUs using Cuda. In: Proc. 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2, pp. 79–84. ACM (2009)
Weyhausen, A.: Numerical algorithms of general relativity for heterogeneous computing environments. Diplomarbeit, Universität Jena, Physics Dept. (2010)
Maruyama, N., Nomura, T., Sato, K., Matsuoka, S.: Physis: An implicitly parallel programming model for stencil computations on large-scale GPU-accelerated supercomputers. In: Supercomputing. IEEE (2011)
Holewinski, J., Pouchet, L.N., Sadayappan, P.: High-performance code generation for stencil computations on GPU architectures. In: Proceedings of the 26th ACM International Conference on Supercomputing (2012)
Williams, S., Kalamkar, D., Singh, A., Deshpande, A., Straalen, B.V., Smelyanskiy, M., Almgren, A., Dubey, P., Shalf, J., Oliker, L.: Optimization of geometric multigrid for emerging multi- and manycore processors. In: Supercomputing. IEEE (2012)
Zhang, Y., Mueller, F.: Auto-generation and auto-tuning of 3D stencil codes on GPU clusters. In: Proc. 10th Int. Symp. Code Gen. Optim., San Jose, CA (2012)
Zumbusch, G.: Tuning a finite difference computation for parallel vector processors. In: 11th Int. Symp. Parallel and Distrib. Comput. CPS, pp. 63–70. IEEE (2012)
Song, Y., Li, Z.: New tiling techniques to improve cache temporal locality. In: Proc. ACM SIGPLAN Conf. Prog. Lang. Design Impl., Atlanta, pp. 215–228 (1999)
McCalpin, J., Wonnacott, D.: Time skewing: A value-based approach to optimizing for memory locality. Technical Report DCS-TR-379, Rutgers Univ. (1999)
Rivera, G., Tseng, C.: Tiling optimizations for 3D scientific computations. In: Supercomputing (2000)
Weiß, C.: Data Locality Optimizations for Multigrid Methods on Structured Grids. PhD thesis, TU München (2001)
Stürmer, M., Treibig, J., Rüde, U.: Optimising a 3D multigrid algorithm for the IA-64 architecture. Int. J. Computational Science and Engineering 4, 29–35 (2008)
Wellein, G., Hager, G., Zeiser, T., Wittmann, M., Fehske, H.: Efficient temporal blocking for stencil computations by multicore-aware wavefront parallelization. In: Int. Comput. Soft. and Applications Conf. (COMPSAC), pp. 579–586 (2009)
Nguyen, A., Satish, N., Chhugani, J., Kim, C., Dubey, P.: 3.5-D blocking optimization for stencil computations on modern CPUs and GPUs. In: Supercomputing. IEEE (2010)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Zumbusch, G. (2013). Vectorized Higher Order Finite Difference Kernels. In: Manninen, P., Öster, P. (eds) Applied Parallel and Scientific Computing. PARA 2012. Lecture Notes in Computer Science, vol 7782. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-36803-5_25
Download citation
DOI: https://doi.org/10.1007/978-3-642-36803-5_25
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-36802-8
Online ISBN: 978-3-642-36803-5
eBook Packages: Computer ScienceComputer Science (R0)