Can I read A script-based autotuning compiler system to generate high-performance CUDA code on EtoBox?

A script-based autotuning compiler system to generate high-performance CUDA code by Malik Khan; Protonu Basu; Gabe Rudy; Mary Hall; Chun Chen; Jacqueline Chame is a Computer Science article available to read on EtoBox.

What is A script-based autotuning compiler system to generate high-performance CUDA code about?

This article presents a novel compiler framework for CUDA code generation. The compiler structure is designed to support __autotuning__ , which employs empirical techniques to evaluate a set of alternative mappings of computation kernels and select the mapping that obtains the best performance. This article introduces a __Transformation Strategy Generator__ , a meta-optimizer that generates a set of __transformation recipes__ , which are descriptions of the mapping of the sequential code to parallel CUDA code. These recipes comprise a search space of possible implementations. This system achieves performance comparable and sometimes better than manually tuned libraries and exceeds the performance of a state-of-the-art GPU compiler.

Who reads A script-based autotuning compiler system to generate high-performance CUDA code?

Author

Malik Khan; Protonu Basu; Gabe Rudy; Mary Hall; Chun Chen; Jacqueline Chame

Publisher

Association for Computing Machinery; Association for Computing Machinery (ACM) (ISSN 1544-3566)

Published

2013

Language

EN

Field

Computer Science (Physical Sciences)