Development of a symbol Manipulation-Oriented Highly Parallel Computer
| Project/Area Number |
01850075
|
|---|
| Research Category |
Grant-in-Aid for Developmental Scientific Research
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
計算機工学
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
SHIBAYAMA Kiyoshi (1990-1991) Kyoto Univ., Eng., Assoc. Prof., 工学部, 助教授 (70127091)
萩原 宏 (1989) 京都大学, 工学部, 教授 (00025818)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
HAGIWARA Hiroshi Ryukoku Univ., Sci. and Tech., Prof., 理工学部, 教授 (00025818)
井上 知子 京都大学, 工学部, 教務職員 (40109145)
新實 治男 京都大学, 工学部, 助手 (40144331)
柴山 潔 京都大学, 工学部, 助教授 (70127091)
|
|---|
| Project Period (FY) |
1989 – 1991
|
| Project Status |
Completed (Fiscal Year 1991)
|
| Budget Amount *help |
¥30,000,000 (Direct Cost: ¥30,000,000)
Fiscal Year 1991: ¥3,100,000 (Direct Cost: ¥3,100,000)
Fiscal Year 1990: ¥9,800,000 (Direct Cost: ¥9,800,000)
Fiscal Year 1989: ¥17,100,000 (Direct Cost: ¥17,100,000)
|
|---|
| Keywords | Parallel Computer / Symbol Manipulation / Symbol Manipulation Language / Highly Parallel Computer / Non-Numeric Processing |
|---|
| Research Abstract |
In this research, we develop the architecture of a logic programming language-oriented highly parallel computer, called KPR, which can execute a program written in a parallel logic programming language KPR-L.
The KPR is controlled on the basis of a new execution model named "Parallel Reduction (PR) model", where a logical process is allocated to a node of an AND/OR process graph and the resultant reduction (folding / unfolding) of this process graph is executed in parallel. This execution model is mainly implemented by the following processes : (i) "Stream-process" for realizing the AND-parallelism by a stream-pipeline processing method ; (ii) "Or-processing for implementing OR-parallelism of a logic program.
KPR is a heterogeneous-function distributed-processing system, where each process is executed on the specialized processor : (i) ARP(And Reduction Processor) for implementing a Stream-parallel processing strategy ; (ii) ORP (Or Reduction Processor) for executing four sets of unification operations in parallel. We have developed a prototype system of KPR which is composed of five units.
|
Report
(4 results)
Research Products
(15 results)
