System-wide profiling and optimization with virtual machines

Shih Hao Hung, Tei Wei Kuo, Chi Sheng Shih, ChiaHeng Tu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

13 Citations (Scopus)

Abstract

Simulation is a common approach for assisting system design and optimization. For system-wide optimization, energy and computational resources are often the two most critical limitations. Modeling energy-states of each hardware component and time spent in each state is needed for accurate energy and performance prediction. Tracking software execution in a realistic operating environment with properly modeled input/output is key to accurate prediction. However, the conventional approaches can have difficulties in practice. First, for a complex system such as an Android smartphone, building a cycle-accurate simulation environment is no easy task. Secondly, for I/O-intensive applications, a slow simulation would significantly alter the application behavior and change its performance profile. Thirdly, conventional software profiling tools generally do not work on simulators, which makes it difficult for performance analysis of complicated software, e.g., Java applications executed by the Dalvik virtual machine. Recently, virtual machine technologies are widely used to emulate a variety of computer systems. While virtual machines do not model the hardware components in the emulated system, we can ease the effort of building a simulation environment by leveraging the infrastructure of virtual machines and adding performance and power models. Moreover, multiple sets of the performance and energy models can be selectively used to verify if the speed of the simulated system impacts the software behavior. Finally, performance monitoring facilities can be integrated to work with profiling tools. We believe this approach should help overcome the aforementioned difficulties. We have prototyped a framework and our case studies showed that the information provided by our tools are useful for software optimization and system design for Android smartphones.

Original languageEnglish
Title of host publicationASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference
Pages395-400
Number of pages6
DOIs
Publication statusPublished - 2012
Event17th Asia and South Pacific Design Automation Conference, ASP-DAC 2012 - Sydney, NSW, Australia
Duration: 2012 Jan 302012 Feb 2

Other

Other17th Asia and South Pacific Design Automation Conference, ASP-DAC 2012
CountryAustralia
CitySydney, NSW
Period12-01-3012-02-02

Fingerprint

Smartphones
Systems analysis
Electron energy levels
Computer hardware
Large scale systems
Computer systems
Simulators
Hardware
Virtual machine
Monitoring
Android (operating system)

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design

Cite this

Hung, S. H., Kuo, T. W., Shih, C. S., & Tu, C. (2012). System-wide profiling and optimization with virtual machines. In ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference (pp. 395-400). [6164980] https://doi.org/10.1109/ASPDAC.2012.6164980
Hung, Shih Hao ; Kuo, Tei Wei ; Shih, Chi Sheng ; Tu, ChiaHeng. / System-wide profiling and optimization with virtual machines. ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference. 2012. pp. 395-400
@inproceedings{7c99222d8ecd4b68904a7102391b1bc3,
title = "System-wide profiling and optimization with virtual machines",
abstract = "Simulation is a common approach for assisting system design and optimization. For system-wide optimization, energy and computational resources are often the two most critical limitations. Modeling energy-states of each hardware component and time spent in each state is needed for accurate energy and performance prediction. Tracking software execution in a realistic operating environment with properly modeled input/output is key to accurate prediction. However, the conventional approaches can have difficulties in practice. First, for a complex system such as an Android smartphone, building a cycle-accurate simulation environment is no easy task. Secondly, for I/O-intensive applications, a slow simulation would significantly alter the application behavior and change its performance profile. Thirdly, conventional software profiling tools generally do not work on simulators, which makes it difficult for performance analysis of complicated software, e.g., Java applications executed by the Dalvik virtual machine. Recently, virtual machine technologies are widely used to emulate a variety of computer systems. While virtual machines do not model the hardware components in the emulated system, we can ease the effort of building a simulation environment by leveraging the infrastructure of virtual machines and adding performance and power models. Moreover, multiple sets of the performance and energy models can be selectively used to verify if the speed of the simulated system impacts the software behavior. Finally, performance monitoring facilities can be integrated to work with profiling tools. We believe this approach should help overcome the aforementioned difficulties. We have prototyped a framework and our case studies showed that the information provided by our tools are useful for software optimization and system design for Android smartphones.",
author = "Hung, {Shih Hao} and Kuo, {Tei Wei} and Shih, {Chi Sheng} and ChiaHeng Tu",
year = "2012",
doi = "10.1109/ASPDAC.2012.6164980",
language = "English",
isbn = "9781467307727",
pages = "395--400",
booktitle = "ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference",

}

Hung, SH, Kuo, TW, Shih, CS & Tu, C 2012, System-wide profiling and optimization with virtual machines. in ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference., 6164980, pp. 395-400, 17th Asia and South Pacific Design Automation Conference, ASP-DAC 2012, Sydney, NSW, Australia, 12-01-30. https://doi.org/10.1109/ASPDAC.2012.6164980

System-wide profiling and optimization with virtual machines. / Hung, Shih Hao; Kuo, Tei Wei; Shih, Chi Sheng; Tu, ChiaHeng.

ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference. 2012. p. 395-400 6164980.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - System-wide profiling and optimization with virtual machines

AU - Hung, Shih Hao

AU - Kuo, Tei Wei

AU - Shih, Chi Sheng

AU - Tu, ChiaHeng

PY - 2012

Y1 - 2012

N2 - Simulation is a common approach for assisting system design and optimization. For system-wide optimization, energy and computational resources are often the two most critical limitations. Modeling energy-states of each hardware component and time spent in each state is needed for accurate energy and performance prediction. Tracking software execution in a realistic operating environment with properly modeled input/output is key to accurate prediction. However, the conventional approaches can have difficulties in practice. First, for a complex system such as an Android smartphone, building a cycle-accurate simulation environment is no easy task. Secondly, for I/O-intensive applications, a slow simulation would significantly alter the application behavior and change its performance profile. Thirdly, conventional software profiling tools generally do not work on simulators, which makes it difficult for performance analysis of complicated software, e.g., Java applications executed by the Dalvik virtual machine. Recently, virtual machine technologies are widely used to emulate a variety of computer systems. While virtual machines do not model the hardware components in the emulated system, we can ease the effort of building a simulation environment by leveraging the infrastructure of virtual machines and adding performance and power models. Moreover, multiple sets of the performance and energy models can be selectively used to verify if the speed of the simulated system impacts the software behavior. Finally, performance monitoring facilities can be integrated to work with profiling tools. We believe this approach should help overcome the aforementioned difficulties. We have prototyped a framework and our case studies showed that the information provided by our tools are useful for software optimization and system design for Android smartphones.

AB - Simulation is a common approach for assisting system design and optimization. For system-wide optimization, energy and computational resources are often the two most critical limitations. Modeling energy-states of each hardware component and time spent in each state is needed for accurate energy and performance prediction. Tracking software execution in a realistic operating environment with properly modeled input/output is key to accurate prediction. However, the conventional approaches can have difficulties in practice. First, for a complex system such as an Android smartphone, building a cycle-accurate simulation environment is no easy task. Secondly, for I/O-intensive applications, a slow simulation would significantly alter the application behavior and change its performance profile. Thirdly, conventional software profiling tools generally do not work on simulators, which makes it difficult for performance analysis of complicated software, e.g., Java applications executed by the Dalvik virtual machine. Recently, virtual machine technologies are widely used to emulate a variety of computer systems. While virtual machines do not model the hardware components in the emulated system, we can ease the effort of building a simulation environment by leveraging the infrastructure of virtual machines and adding performance and power models. Moreover, multiple sets of the performance and energy models can be selectively used to verify if the speed of the simulated system impacts the software behavior. Finally, performance monitoring facilities can be integrated to work with profiling tools. We believe this approach should help overcome the aforementioned difficulties. We have prototyped a framework and our case studies showed that the information provided by our tools are useful for software optimization and system design for Android smartphones.

UR - http://www.scopus.com/inward/record.url?scp=84859983399&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84859983399&partnerID=8YFLogxK

U2 - 10.1109/ASPDAC.2012.6164980

DO - 10.1109/ASPDAC.2012.6164980

M3 - Conference contribution

SN - 9781467307727

SP - 395

EP - 400

BT - ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference

ER -

Hung SH, Kuo TW, Shih CS, Tu C. System-wide profiling and optimization with virtual machines. In ASP-DAC 2012 - 17th Asia and South Pacific Design Automation Conference. 2012. p. 395-400. 6164980 https://doi.org/10.1109/ASPDAC.2012.6164980