Loop transformations leveraging hardware prefetching
Savvas Sioutas, Sander Stuijk, Henk Corporaal et al. · International Symposium on Code Generation and Optimization (CGO) · 2018
Memory-bound applications heavily depend on the bandwidth of the system in order to achieve high performance. Improving temporal and/or spatial locality through loop transformations is a common way of mitigating this dependency. However, choosing the right combination of optimizations is not a trivial task, due to the fact that most of them alter the memory access pattern of the application and as a result interfere with the efficiency of the hardware prefetching mechanisms present in modern architectures. We propose an optimization algorithm that analytically classifies an algorithmic description of a loop nest in order to decide whether it should be optimized stressing its temporal or spatial locality, while also taking hardware prefetching into account. We implement our technique as a tool to be used with the Halide compiler and test it on a variety of benchmarks. We find an average performance improvement of over 40% compared to previous analytical models targeting the Halide language and compiler.
Cited by 14 works. doi:10.1145/3168823
Authors
References
- A data locality optimizing algorithm
- Automated empirical optimizations of software and the ATLAS project
- The cache performance and optimizations of blocked algorithms
- Automatically Tuned Linear Algebra Software
- Halide
- Automatically Tuned Linear Algebra Software
- Tile size selection using cache organization and data layout
- Self-Adapting Linear Algebra Algorithms and Software
- OpenTuner
- Impact of modern memory subsystems on cache optimizations for stencil computations