Advanced search
1 file | 228.94 KB Add to list

Control flow modeling in statistical simulation for accurate and efficient processor design studies

Author
Organization
Abstract
Designing a new microprocessor is extremely time-consuming. One of the contributing reasons is that computer designers rely heavily on detailed architectural simulations, which are very time-consuming. Recent work has focused on statistical simulation to address this issue. The basic idea of statistical simulation is to measure characteristics during program execution, generate a synthetic trace with those characteristics and then simulate the synthetic trace. The statistically generated synthetic trace is orders of magnitude smaller than the original program sequence and hence results in significantly faster simulation. This paper makes the following contributions to the statistical simulation methodology. First, we propose the use of a statistical flow graph to characterize the control flow of a program execution. Second, we model delayed update of branch predictors while profiling program execution characteristics. Experimental results show that statistical simulation using this improved control flow modeling attains significantly better accuracy than the previously proposed HLS system. We evaluate both the absolute and the relative accuracy of our approach for power/performance modeling of superscalar microarchitectures. The results show that our statistical simulation framework can be used to efficiently explore processor design spaces.
Keywords
PERFORMANCE

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 228.94 KB

Citation

Please use this url to cite or link to this publication:

MLA
Eeckhout, Lieven, Robert H Jr Bell, Bastiaan Stougie, et al. “Control Flow Modeling in Statistical Simulation for Accurate and Efficient Processor Design Studies.” Conference Proceedings - Annual International Symposium on Computer Architecture. Los Alamitos, CA, USA: IEEE Computer Society, 2004. 350–361. Print.
APA
Eeckhout, L., Bell, R. H. J., Stougie, B., De Bosschere, K., & John, L. K. (2004). Control flow modeling in statistical simulation for accurate and efficient processor design studies. CONFERENCE PROCEEDINGS - ANNUAL INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE (pp. 350–361). Presented at the 31st Annual International Symposium on Computer Architecture, Los Alamitos, CA, USA: IEEE Computer Society.
Chicago author-date
Eeckhout, Lieven, Robert H Jr Bell, Bastiaan Stougie, Koen De Bosschere, and Lizy K John. 2004. “Control Flow Modeling in Statistical Simulation for Accurate and Efficient Processor Design Studies.” In Conference Proceedings - Annual International Symposium on Computer Architecture, 350–361. Los Alamitos, CA, USA: IEEE Computer Society.
Chicago author-date (all authors)
Eeckhout, Lieven, Robert H Jr Bell, Bastiaan Stougie, Koen De Bosschere, and Lizy K John. 2004. “Control Flow Modeling in Statistical Simulation for Accurate and Efficient Processor Design Studies.” In Conference Proceedings - Annual International Symposium on Computer Architecture, 350–361. Los Alamitos, CA, USA: IEEE Computer Society.
Vancouver
1.
Eeckhout L, Bell RHJ, Stougie B, De Bosschere K, John LK. Control flow modeling in statistical simulation for accurate and efficient processor design studies. CONFERENCE PROCEEDINGS - ANNUAL INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE. Los Alamitos, CA, USA: IEEE Computer Society; 2004. p. 350–61.
IEEE
[1]
L. Eeckhout, R. H. J. Bell, B. Stougie, K. De Bosschere, and L. K. John, “Control flow modeling in statistical simulation for accurate and efficient processor design studies,” in CONFERENCE PROCEEDINGS - ANNUAL INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE, Munich, Germany, 2004, pp. 350–361.
@inproceedings{404140,
  abstract     = {Designing a new microprocessor is extremely time-consuming. One of the contributing reasons is that computer designers rely heavily on detailed architectural simulations, which are very time-consuming. Recent work has focused on statistical simulation to address this issue. The basic idea of statistical simulation is to measure characteristics during program execution, generate a synthetic trace with those characteristics and then simulate the synthetic trace. The statistically generated synthetic trace is orders of magnitude smaller than the original program sequence and hence results in significantly faster simulation.
This paper makes the following contributions to the statistical simulation methodology. First, we propose the use of a statistical flow graph to characterize the control flow of a program execution. Second, we model delayed update of branch predictors while profiling program execution characteristics. Experimental results show that statistical simulation using this improved control flow modeling attains significantly better accuracy than the previously proposed HLS system. We evaluate both the absolute and the relative accuracy of our approach for power/performance modeling of superscalar microarchitectures. The results show that our statistical simulation framework can be used to efficiently explore processor design spaces.},
  author       = {Eeckhout, Lieven and Bell, Robert H Jr and Stougie, Bastiaan and De Bosschere, Koen and John, Lizy K},
  booktitle    = {CONFERENCE PROCEEDINGS - ANNUAL INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE},
  isbn         = {0-7695-2143-6},
  issn         = {1063-6897},
  keywords     = {PERFORMANCE},
  language     = {eng},
  location     = {Munich, Germany},
  pages        = {350--361},
  publisher    = {IEEE Computer Society},
  title        = {Control flow modeling in statistical simulation for accurate and efficient processor design studies},
  year         = {2004},
}

Web of Science
Times cited: