3.5-D Blocking Optimization for Stencil Computations on Modern CPUs and GPUs

Citations Over TimeTop 1% of 2010 papers

Abstract

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.

Related Papers

• → Data Layout Transformation for Stencil Computations on Short-Vector SIMD Architectures(2011)112 cited
• → Multi-level spatial and temporal tiling for efficient HPC stencil computation on many-core processors with large shared caches(2017)15 cited
• → Memory Bandwidth: The True Bottleneck of SIMD Multimedia Performance on a Superscalar Processor(2001)14 cited
• → Efficient Emulation of Floating-Point Arithmetic on Fixed-Point SIMD Processors(2016)6 cited
• → Hybrid system, a new low cost parallel cluster(2003)4 cited