Compiling techniques for first-order linear recurrences on a vector computer |
| |
Authors: | Yoshikazu Tanaka Kyouko Iwasawa Yukio Umetani Shizuo Gotou |
| |
Affiliation: | (1) Central Research Laboratory, Hitachi, Ltd., Higashikoigakubo, Kokubunji-shi, 185 Tokyo, Japan;(2) Software Works, Hitachi, Ltd., Shinanomachi, Totsuka-ku, Yokohama-shi, 244 Kanagawa, Japan |
| |
Abstract: | Linear recurrences are the most important class of nonvectorizable problems in typical scientific/engineering calculations. This work discusses high-performance methods for solving first-order linear recurrences on a vector computer, investigates automatic transformations, and develops compiling techniques for first-order linear recurrence problems. The results show that the improved vector code generated by the vectorizing compiler on the HITAC S-820 supercomputer runs at the rate of 150 MFLOPS (million floating operations per second) for moderate loop lengths (>1000) and over 200 MFLOPS for long loop lengths (> 10000). Also, overall performance improvements of 69% in the 14 Lawrence Livermore Loops and 25 % in the 24 Lawrence Livermore Loops, as measured by the harmonic mean, are attained. |
| |
Keywords: | First-order linear recurrence interstatement recurrence vectorizing compiler program transformation macro-vector instruction data dependence |
本文献已被 SpringerLink 等数据库收录! |
|