Beating C in scientific computing applications

Abstract

This paper presents an ongoing research on the behavior and performance of Lisp with respect to C in the context of scientific numerical computing. Several simple image processing algorithms are used to evaluate the performance of pixel access and arithmetic operations in both languages. We demonstrate that the behavior of equivalent Lisp and C code is similar with respect to the choice of data structures and types, and also to external parameters such as hardware optimization. We further demonstrate that properly typed and optimized Lisp code runs as fast as the equivalent C code, or even faster in some cases.