论文信息 - Compiling Techniques for Supercomputers

Compiling Techniques for Supercomputers

Summary Although the architecture of Japanese supercomputers is of the CRAY-1 type, important architectural features have been added to improve their performance, and these have a significant influence on the software. After giving a brief historical survey of programming languages for vector/parallel computers in Japan, we describe compiling techniques for supercomputers in Japan.

Masaaki Shimasaki

[1] Olaf M. Lubeck,et al. The birth of the second generation: the Hitachi S-820/80 , 1988, Proceedings. SUPERCOMPUTING '88.

[2] M. Shimasaki. On techniques in vectorizing compilers and optimizing program transformations for supercomputers , 1987 .

[3] Richard M. Russell,et al. The CRAY-1 computer system , 1978, CACM.

[4] David A. Padua,et al. Advanced compiler optimizations for supercomputers , 1986, CACM.

[5] Yoshitoshi Kunieda,et al. V-Pascal: An automatic vectorizing compiler for Pascal with no language extensions , 1988, Supercomputing '88.

[6] Y. Tanaka,et al. Advanced vectorization techniques for supercomputers , 1987 .

[7] Jack J. Dongarra,et al. Squeezing the most out of an algorithm in CRAY FORTRAN , 1984, ACM Trans. Math. Softw..

[8] Ken Kennedy,et al. Automatic loop interchange , 2004, SIGP.

[9] Yukio Umetani,et al. Compiling techniques for first-order linear recurrences on a vector computer , 1988, Supercomputing '88.