Communication-Free Alignment for Array References with Linear Subscripts in Three Loop Index Variables or Quadratic Subscripts

  • Authors:

  • Weng-Long Chang;Chih-Ping Chu;Jia-Hwa Wu

  • Affiliations:

  • Department of Management Information System, Kung Shan Technology University, Tainan, Taiwan 701, R.O.C.;Department of Computer Science and Information Engineering, National Cheng Kung University, Tainan, Taiwan 701, R.O.C. chucp@csie.ncku.edu.tw;Department of Computer Science and Information Engineering, National Cheng Kung University, Tainan, Taiwan 701, R.O.C.

  • Venue:
  • The Journal of Supercomputing
  • Year:
  • 2001

Quantified Score

Hi-index 0.00

Visualization

Abstract

Bau et al. proposed an efficient and precise data alignment method to ascertain whether there is communication-free alignment of array reference function with linear subscripts in one loop index variable. Chu et al. presented an efficient and precise data alignment method to determine whether there is communication-free alignment of array reference function with linear subscripts in two loop index variables or quadratic subscripts (ai^2+bi+d). However, for array reference function with linear subscripts in three loop index variables or quadratic subscripts (ai^2+bi+cj+d), their methods cannot be applied. In this paper, we propose two new alignment functions for loop iteration space and array elements. The new alignment functions can be applied towards checking whether there is communication-free alignment of array reference function with linear subscripts in three loop index variables or quadratic subscripts. Experiments with benchmarks taken from Parallel loop and Vector loop showed that among the 7 nested loops tested, three of them had their data alignment improved by the method proposed.