Exploiting parallelism in cyclic pipeline computer with an optimizing compiler 最適化コンパイラによる循環パイプライン・コンピューターにおける並列化の研究
Access this Article
Search this Article
Author
Bibliographic Information
- Title
-
Exploiting parallelism in cyclic pipeline computer with an optimizing compiler
- Other Title
-
最適化コンパイラによる循環パイプライン・コンピューターにおける並列化の研究
- Author
-
佐藤, 三久
- Author(Another name)
-
サトウ, ミツヒサ
- University
-
東京大学
- Types of degree
-
理学博士
- Grant ID
-
乙第10044号
- Degree year
-
1991-02-21
Note and Description
博士論文
Table of Contents
- ABSTRACT / (0003.jp2)
- Contents / p1 (0005.jp2)
- 1 Introduction / p7 (0011.jp2)
- 1.1 Cyclic Pipeline Computer / p7 (0011.jp2)
- 1.2 FLATS2 / p8 (0012.jp2)
- 1.3 FLATS2 FORTRAN Compiler / p8 (0012.jp2)
- 1.4 Parallel Programming Model / p9 (0013.jp2)
- 1.5 FLATS2 project / p9 (0013.jp2)
- 1.6 Scope of Study in This Thesis / p10 (0014.jp2)
- 2 Highly pipelined Model / p11 (0015.jp2)
- 2.1 Models of Highly Pipelined Machines / p11 (0015.jp2)
- 2.2 Parallel Programming Model / p15 (0019.jp2)
- 2.3 Simulation / p18 (0022.jp2)
- 2.4 Discussion / p25 (0029.jp2)
- 3 A Cyclic Pipeline Computer,FLATS2 / p28 (0032.jp2)
- 3.1 The architecture of FLATS2 / p28 (0032.jp2)
- 3.2 Programming on FLATS2 / p33 (0037.jp2)
- 3.3 Programming Environment / p35 (0039.jp2)
- 4 FLATS2 FORTRAN compiler / p37 (0041.jp2)
- 4.1 Overview / p37 (0041.jp2)
- 4.2 Code Expansion Phase / p40 (0044.jp2)
- 4.3 Translation into Static Single Assignment Form / p41 (0045.jp2)
- 4.4 Redundancy Elimination / p44 (0048.jp2)
- 4.5 Code Reconstruction / p49 (0053.jp2)
- 4.6 Register Allocation / p52 (0056.jp2)
- 5 Loop optimization with BL Addressing / p62 (0066.jp2)
- 5.1 Addressing mode for Numerical Computation / p62 (0066.jp2)
- 5.2 Optimal Induction Variable Elimination / p64 (0068.jp2)
- 5.3 BL Code Generation / p74 (0078.jp2)
- 6 Experiments on FLATS2 / p78 (0082.jp2)
- 6.1 The performance of FLATS2 / p78 (0082.jp2)
- 6.2 The Experiments on the FLATS2 Pipeline / p81 (0085.jp2)
- 7 Summary and Conclusions / p88 (0092.jp2)
- A RUN-TIME CHECKING IN LISP BY INTEGRATING MEMORY AD_DRESSING AND RANGE CHECKING / p90 (0094.jp2)
- B Benchmark programs in Parallel FORTRAN / p99 (0103.jp2)