摘要
科学计算中的许多领域都需要快速而精确地计算超越函数,即exp、log、sin、tan等此类函数。本文采用表驱动算法,结合IA-64体系结构特点,在GCC中优化实现了指数函数(exp),提高了GCC编译器在IA-64系统上的浮点性能,为在IA-64和其它平台上高效实现所有超越函数打下了基础。
The fast and accurate evaluation of transcendental functions (e. g. exp, log, sin, tan) is very important in the field of scientific computing. This paper presents the table-driven implementation of the exponential function in GCC by exploiting some key features of the IA-64 architecture. This work not only improves the floating-point performance of GCC on IA-64, but also guides the implementation of other transcendental functions both on IA-64 and other platforms.
出处
《计算机工程与科学》
CSCD
2007年第5期77-80,共4页
Computer Engineering & Science
基金
国家863计划资助项目(2004AA1Z2210)