Expression Templates Revisited: A Performance Analysis of the Current ET Methodology

Abstract

In the last decade, Expression Templates (ET) have gained a reputation as an efficient performance optimization tool for C++ codes. This reputation builds on several ET-based linear algebra frameworks focused on combining both elegant and high-performance C++ code. However, on closer examination the assumption that ETs are a performance optimization technique cannot be maintained. In this paper we demonstrate and explain the inability of current ET-based frameworks to deliver high performance for dense and sparse linear algebra operations, and introduce a new "smart" ET implementation that truly allows the combination of high performance code with the elegance and maintainability of a domain-specific language.

16 pages, 7 figures

Cited by in corpus (19)

• Practical Sparse Matrices in C++ with Hybrid Storage and Template-Based Expression Optimisation
• The LifeV library: engineering mathematics beyond the proof of concept
• An asynchronous and task-based implementation of Peridynamics utilizing HPX -- the C++ standard library for parallelism and concurrency
• Asynchronous Execution of Python Code on Task Based Runtime Systems
• Best practices for HPM-assisted performance engineering on modern multicore processors
• Yet Another Tensor Toolbox for discontinuous Galerkin methods and other applications
• The Linear Algebra Mapping Problem. Current state of linear algebra languages and libraries
• Learning to play the Chess Variant Crazyhouse above World Champion Level with Deep Neural Networks and Human Data
• Large scale simulation of pressure induced phase-field fracture propagation using Utopia
• The Generalized Matrix Chain Algorithm
• Eigen-AD: Algorithmic Differentiation of the Eigen Library
• Particle-based and Meshless Methods with Aboria
• Worldtube excision method for intermediate-mass-ratio inspirals: scalar-field model in 3+1 dimensions
• Automatic Generation of Efficient Linear Algebra Programs
• Scheduling optimization of parallel linear algebra algorithms using Supervised Learning
• Worldtube excision method for intermediate-mass-ratio inspirals: self-consistent evolution in a scalar-charge model
• Relieving scale disparity in binary black hole simulations
• FDBB: Fluid Dynamics Building Blocks
• Model-guided Performance Analysis of the Sparse Matrix-Matrix Multiplication