3.5-D Blocking Optimization for Stencil Computations on Modern CPUs and GPUs
Anthony D. Nguyen, Nadathur Satish, Jatin Chhugani et al. · 2010
Stencil computation sweeps over a spatial grid over multiple time steps to perform nearest-neighbor computations. The bandwidth-to-compute requirement for a large class of stencil kernels is very high, and their performance is bound by the available memory bandwidth. Since memory bandwidth grows slower than compute, the performance of stencil kernels will not scale with increasing compute density. We present a novel 3.5D-blocking algorithm that performs 2.5D-spatial and temporal blocking of the input grid into on-chip memory for both CPUs and GPUs. The resultant algorithm is amenable to both thread- level and data-level parallelism, and scales near-linearly with the SIMD width and multiple-cores. Our performance numbers are faster or comparable to state-of-the-art-stencil implementations on CPUs and GPUs. Our implementation of 7-point-stencil is 1.5X-faster on CPUs, and 1.8X faster on GPUs for single- precision floating point inputs than previously reported numbers. For Lattice Boltzmann methods, the corresponding speedup number on CPUs is 2.1X.
Cited by 290 works. doi:10.1109/sc.2010.2
Authors
References
- Adaptive mesh refinement for hyperbolic partial differential equations
- Adaptive Mesh Refinement for Hyperbolic Partial Differential Equations
- Algorithms for scalable synchronization on shared-memory multiprocessors
- Synthesis and evaluation of linear motion transitions
- Larrabee
- Stencil computation optimization and auto-tuning on state-of-the-art multicore architectures
- 3D finite difference computation on GPUs using CUDA
- Salinity-driven Thermocline Transients in a Wind- and Thermohaline-forced Isopycnic Coordinate Model of the North Atlantic
- Implicit and explicit optimizations for stencil computations
- Tiling Optimizations for 3D Scientific Computations