Annotation-based empirical performance tuning using Orio

TLDR

Many scientific applications spend significant time tuning code for specific high‑performance architectures, yet intrusive techniques can make code hard to reverse, maintain, or port. Orio is an extensible annotation‑based empirical tuning system designed to improve both performance and productivity. It lets developers insert structured comment annotations that trigger low‑level optimizations, supports both architecture‑independent and architecture‑specific tuning, generates multiple tuned versions for empirical evaluation, and integrates Pluto to produce efficient OpenMP parallel code. Experiments on dense array and sparse matrix kernels demonstrate that Orio can generate efficient tuned versions.

Abstract

For many scientific applications, significant time is spent in tuning codes for a particular high-performance architecture. Tuning approaches range from the relatively nonintrusive (e.g., by using compiler options) to extensive code modifications that attempt to exploit specific architecture features. Intrusive techniques often result in code changes that are not easily reversible, and can negatively impact readability, maintainability, and performance on different architectures. We introduce an extensible annotation-based empirical tuning system called Orio that is aimed at improving both performance and productivity. It allows software developers to insert annotations in the form of structured comments into their source code to trigger a number of low-level performance optimizations on a specified code fragment. To maximize the performance tuning opportunities, the annotation processing infrastructure is designed to support both architecture-independent and architecture-specific code optimizations. Given the annotated code as input, Orio generates many tuned versions of the same operation and empirically evaluates the alternatives to select the best performing version for production use. We have also enabled the use of the Pluto automatic parallelization tool in conjunction with Orio to generate efficient OpenMP-based parallel code. We describe our experimental results involving a number of computational kernels, including dense array and sparse matrix operations.