Advanced search
1 file | 2.71 MB

Selecting representative benchmark inputs for exploring microprocessor design spaces

Maximilien Breughe (UGent) and Lieven Eeckhout (UGent)
Author
Organization
Abstract
The design process of a microprocessor requires representative workloads to steer the search process toward an optimum design point for the target application domain. However, considering a broad set of workloads to cover the large space of potential workloads is infeasible given how time-consuming design space exploration typically is. Hence, it is crucial to select a small yet representative set of workloads, which leads to a shorter design cycle while yielding a (near) optimal design. Prior work has mostly looked into selecting representative benchmarks; however, limited attention was given to the selection of benchmark inputs and how this affects workload representativeness during design space exploration. Using a set of 1,000 inputs for a number of embedded benchmarks and a design space with around 1,700 design points, we find that selecting a single or three random input(s) per benchmark potentially (in a worst-case scenario) leads to a suboptimal design that is 56% and 33% off, on average, relative to the optimal design in our design space in terms of Energy-Delay Product (EDP). We then propose and evaluate a number of methods for selecting representative inputs and show that we can find the optimum design point with as few as three inputs.
Keywords
input selection, workload selection, SUITE, Processor design space exploration

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 2.71 MB

Citation

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

Chicago
Breughe, Maximilien, and Lieven Eeckhout. 2013. “Selecting Representative Benchmark Inputs for Exploring Microprocessor Design Spaces.” Acm Transactions on Architecture and Code Optimization 10 (4): 1–24.
APA
Breughe, M., & Eeckhout, L. (2013). Selecting representative benchmark inputs for exploring microprocessor design spaces. ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION, 10(4), 1–24.
Vancouver
1.
Breughe M, Eeckhout L. Selecting representative benchmark inputs for exploring microprocessor design spaces. ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION. ACM; 2013;10(4):1–24.
MLA
Breughe, Maximilien, and Lieven Eeckhout. “Selecting Representative Benchmark Inputs for Exploring Microprocessor Design Spaces.” ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION 10.4 (2013): 1–24. Print.
@article{4375101,
  abstract     = {The design process of a microprocessor requires representative workloads to steer the search process toward an optimum design point for the target application domain. However, considering a broad set of workloads to cover the large space of potential workloads is infeasible given how time-consuming design space exploration typically is. Hence, it is crucial to select a small yet representative set of workloads, which leads to a shorter design cycle while yielding a (near) optimal design. Prior work has mostly looked into selecting representative benchmarks; however, limited attention was given to the selection of benchmark inputs and how this affects workload representativeness during design space exploration. Using a set of 1,000 inputs for a number of embedded benchmarks and a design space with around 1,700 design points, we find that selecting a single or three random input(s) per benchmark potentially (in a worst-case scenario) leads to a suboptimal design that is 56\% and 33\% off, on average, relative to the optimal design in our design space in terms of Energy-Delay Product (EDP). We then propose and evaluate a number of methods for selecting representative inputs and show that we can find the optimum design point with as few as three inputs.},
  articleno    = {37},
  author       = {Breughe, Maximilien and Eeckhout, Lieven},
  issn         = {1544-3566},
  journal      = {ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION},
  language     = {eng},
  number       = {4},
  pages        = {37:1--37:24},
  publisher    = {ACM},
  title        = {Selecting representative benchmark inputs for exploring microprocessor design spaces},
  url          = {http://dx.doi.org/10.1145/2541228.2555294},
  volume       = {10},
  year         = {2013},
}

Altmetric
View in Altmetric
Web of Science
Times cited: