SEESMA: A Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture

Chao-Chin Wu, Der Lin Pean, Jen Pin Su, Jia Rong Wu, Hsuan Woei Huang, Jun Long Huang, Jia Lin Lee, Huey Ting Chua, Cheng Chen

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

We have constructed a simulation environment called SEESMA (a Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture). It is a program-driven simulator consisting of a memory reference generator and a target system simulator. SEESMA aids in the investigation of shared-memory multiprocessor architecture through a user friendly interface. It is a software platform with a modular structure suitable for educational and research purposes. To achieve the design objectives, SEESMA supports the following simulation environment for various important design spaces: (1) two types of processing element architectures; (2) a two-level cache structure with write cache; (3) five types of memory consistency models; (4) five types of cache coherence protocols; (5) three types of directory structures; (6) interconnection networks, and other related issues. Each sub-environment provides many options for investigating the interactions between different options. Meanwhile, we have also developed an X-window interface to specify system architectures and benchmarks in addition to friendly on-line help. We can determine the setup values of architectural parameters in a status region by moving the cursor around the graph of each architecture component. The architecture graph is changed each time the user modifies the simulation options. In addition, SEESMA can automatically collect data from several simulation results and display a comparison using either a bar graph or a curve graph. So far, many design issues and important performance gains for high performance multiprocessor systems can be evaluated and explored using this platform.

Original languageEnglish
Pages (from-to)524-538
Number of pages15
JournalProceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering
Volume22
Issue number4
Publication statusPublished - 1998 Jul 1

Fingerprint

Data storage equipment
Simulators
User interfaces
Computer systems
Network protocols
Processing

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Wu, Chao-Chin ; Pean, Der Lin ; Su, Jen Pin ; Wu, Jia Rong ; Huang, Hsuan Woei ; Huang, Jun Long ; Lee, Jia Lin ; Chua, Huey Ting ; Chen, Cheng. / SEESMA : A Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture. In: Proceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering. 1998 ; Vol. 22, No. 4. pp. 524-538.
@article{0c7be129c1974c8785aa60995fe9c42a,
title = "SEESMA: A Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture",
abstract = "We have constructed a simulation environment called SEESMA (a Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture). It is a program-driven simulator consisting of a memory reference generator and a target system simulator. SEESMA aids in the investigation of shared-memory multiprocessor architecture through a user friendly interface. It is a software platform with a modular structure suitable for educational and research purposes. To achieve the design objectives, SEESMA supports the following simulation environment for various important design spaces: (1) two types of processing element architectures; (2) a two-level cache structure with write cache; (3) five types of memory consistency models; (4) five types of cache coherence protocols; (5) three types of directory structures; (6) interconnection networks, and other related issues. Each sub-environment provides many options for investigating the interactions between different options. Meanwhile, we have also developed an X-window interface to specify system architectures and benchmarks in addition to friendly on-line help. We can determine the setup values of architectural parameters in a status region by moving the cursor around the graph of each architecture component. The architecture graph is changed each time the user modifies the simulation options. In addition, SEESMA can automatically collect data from several simulation results and display a comparison using either a bar graph or a curve graph. So far, many design issues and important performance gains for high performance multiprocessor systems can be evaluated and explored using this platform.",
author = "Chao-Chin Wu and Pean, {Der Lin} and Su, {Jen Pin} and Wu, {Jia Rong} and Huang, {Hsuan Woei} and Huang, {Jun Long} and Lee, {Jia Lin} and Chua, {Huey Ting} and Cheng Chen",
year = "1998",
month = "7",
day = "1",
language = "English",
volume = "22",
pages = "524--538",
journal = "Proceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering",
issn = "0255-6588",
publisher = "National Science Council",
number = "4",

}

SEESMA : A Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture. / Wu, Chao-Chin; Pean, Der Lin; Su, Jen Pin; Wu, Jia Rong; Huang, Hsuan Woei; Huang, Jun Long; Lee, Jia Lin; Chua, Huey Ting; Chen, Cheng.

In: Proceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering, Vol. 22, No. 4, 01.07.1998, p. 524-538.

Research output: Contribution to journalArticle

TY - JOUR

T1 - SEESMA

T2 - A Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture

AU - Wu, Chao-Chin

AU - Pean, Der Lin

AU - Su, Jen Pin

AU - Wu, Jia Rong

AU - Huang, Hsuan Woei

AU - Huang, Jun Long

AU - Lee, Jia Lin

AU - Chua, Huey Ting

AU - Chen, Cheng

PY - 1998/7/1

Y1 - 1998/7/1

N2 - We have constructed a simulation environment called SEESMA (a Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture). It is a program-driven simulator consisting of a memory reference generator and a target system simulator. SEESMA aids in the investigation of shared-memory multiprocessor architecture through a user friendly interface. It is a software platform with a modular structure suitable for educational and research purposes. To achieve the design objectives, SEESMA supports the following simulation environment for various important design spaces: (1) two types of processing element architectures; (2) a two-level cache structure with write cache; (3) five types of memory consistency models; (4) five types of cache coherence protocols; (5) three types of directory structures; (6) interconnection networks, and other related issues. Each sub-environment provides many options for investigating the interactions between different options. Meanwhile, we have also developed an X-window interface to specify system architectures and benchmarks in addition to friendly on-line help. We can determine the setup values of architectural parameters in a status region by moving the cursor around the graph of each architecture component. The architecture graph is changed each time the user modifies the simulation options. In addition, SEESMA can automatically collect data from several simulation results and display a comparison using either a bar graph or a curve graph. So far, many design issues and important performance gains for high performance multiprocessor systems can be evaluated and explored using this platform.

AB - We have constructed a simulation environment called SEESMA (a Simulation and Evaluation Environment for Shared-memory Multiprocessor Architecture). It is a program-driven simulator consisting of a memory reference generator and a target system simulator. SEESMA aids in the investigation of shared-memory multiprocessor architecture through a user friendly interface. It is a software platform with a modular structure suitable for educational and research purposes. To achieve the design objectives, SEESMA supports the following simulation environment for various important design spaces: (1) two types of processing element architectures; (2) a two-level cache structure with write cache; (3) five types of memory consistency models; (4) five types of cache coherence protocols; (5) three types of directory structures; (6) interconnection networks, and other related issues. Each sub-environment provides many options for investigating the interactions between different options. Meanwhile, we have also developed an X-window interface to specify system architectures and benchmarks in addition to friendly on-line help. We can determine the setup values of architectural parameters in a status region by moving the cursor around the graph of each architecture component. The architecture graph is changed each time the user modifies the simulation options. In addition, SEESMA can automatically collect data from several simulation results and display a comparison using either a bar graph or a curve graph. So far, many design issues and important performance gains for high performance multiprocessor systems can be evaluated and explored using this platform.

UR - http://www.scopus.com/inward/record.url?scp=0032114669&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032114669&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0032114669

VL - 22

SP - 524

EP - 538

JO - Proceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering

JF - Proceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering

SN - 0255-6588

IS - 4

ER -