Predictive Modeling in a Polyhedral Optimization Space

Abstract

Abstract-Significant advances in compiler optimization have been made in recent years, enabling many transformations such as tiling, fusion, parallelization and vectorization on imperfectly nested loops. Nevertheless, the problem of finding the best combination of loop transformations remains a major challenge. Polyhedral models for compiler optimization have demonstrated strong potential for enhancing program performance, in particular for compute-intensive applications. But existing static cost models to optimize polyhedral transformations have significant limitations, and iterative compilation has become a very promising alternative to these models to find the most effective transformations. But since the number of polyhedral optimization alternatives can be enormous, it is often impractical to iterate over a significant fraction of the entire space of polyhedrally transformed variants. Recent research has focused on iterating over this search space either with manually-constructed heuristics or with automatic but very expensive search algorithms (e.g., genetic algorithms) that can eventually find good points in the polyhedral space.In this paper, we propose the use of machine learning to address the problem of selecting the best polyhedral optimizations. We show that these models can quickly find high-performance program variants in the polyhedral space, without resorting to extensive empirical search. We introduce models that take as input a characterization of a program based on its dynamic behavior, and predict the performance of aggressive high-level polyhedral transformations that includes tiling, parallelization and vectorization. We allow for a minimal empirical search on the target machine, discovering on average 83% of the searchspace-optimal combinations in at most 5 runs. Our end-to-end framework is validated using numerous benchmarks on two multi-core platforms.

Using graph-based program characterization for predictive modeling

“…This model has been used in recent work [9,13,22] to predict good optimization sequences for various different compilers. We refer to this model as the speedup predictor because it takes as input a program's characterization (P) and an optimization sequence (O), and it outputs the predicted speedup over some baseline for that optimization sequence.…”

“…Using a well-constructed machine learning model to choose optimizations for a specific program has repeatedly been shown to outperform the most aggressive optimization levels in open-source and commercial compilers [6,10,13,15,16,19,20,22,25,30]. However, to use machine learning effectively, it is critical to use expressive features that characterize programs well and that strongly correlate to beneficial optimization sequences for the target program.…”

“…Methods to dynamically characterize a program have consisted of either using performance counters or a technique known as "reactions." Performance counters can be used to collect information, such as cache behavior, functional unit usage, and dynamic instruction mix from the running program [10,15,22]. Reactions is a technique to dynamically characterize a program by selecting specific program transformations to apply to a program and using the resulting speedups to characterize a program's behavior [9,27].…”

Guide Automatic Vectorization by means of Machine Learning: A Case Study of Tensor Contraction Kernels

“…molecule-structure simulation). Moreover, we focus on the Intel Compiler over the GCC * and LLVM * * compilers for two reasons: (1) GCC and LLVM almost invariably fail to optimize the TC kernels, (2) only the Intel Compiler provides a way to force vectorization to vectorize through the command line, thereby bypassing its performance-prediction heuristics. We measure in Sect.…”

“…Using machine learning to predict the best optimization strategy is not a novel idea, and this approach has already proved to be effective in a number of general-purpose datasets [2], [6]. The software characteristics are numbers fed to a machine learning algorithm that outputs the compiler options.…”