Anish Muttreja
Anand Raghunathan
ABSTRACT
Efficient energy and performance estimation of embedded software is a critical part of any system-level design flow. Macromodeling based estimation is an attempt to speed up estimation by exploiting reuse that is inherent in the design process. Macromodeling involves pre-characterizing reusable software components to construct high-level models, which express the execution time or energy consumption of a sub-program as a function of suitable parameters. During simulation, macromodels can be used instead of detailed hardware models, resulting in orders of magnitude simulation speedup. However, in order to realize this potential, significant challenges need to be overcome in both the generation and use of macromodels--- including how to identify the parameters to be used in the macromodel, how to define the template function to which the macromodel is fitted, em etc. This paper presents an automatic methodology to perform characterization-based high-level software macromodeling, which addresses the aforementioned issues. Given a sub-program to be macromodeled for execution time and/or energy consumption, the proposed methodology automates the steps of parameter identification, data collection through detailed simulation, macromodel template selection, and fitting. We propose a novel technique to identify potential macromodel parameters and perform data collection, which draws from the concept of bf data structure serialization used in distributed programming. We utilize bf symbolic regression techniques to concurrently filter out irrelevant macromodel parameters, construct a macromodel function, and derive the optimal coefficient values to minimize fitting error. Experiments with several realistic benchmarks suggest that the proposed methodology improves estimation accuracy and enables wide applicability of macromodeling to complex embedded software, while realizing its potential for estimation speedup. We describe a case study of how macromodeling can be used to rapidly explore algorithm-level energy tradeoffs, for the tt zlib data compression library.
- J. Rabaey and M. Pedram (Editors), Low Power Design Methodologies.hskip 1em plus 0.5em minus 0.4emrelax Kluwer Academic Publishers, Norwell, MA, 1996. Google Scholar
Digital Library
- A. Raghunathan, N. K. Jha, and S. Dey, High-level Power Analysis and Optimization. Kluwer Academic Publishers, Norwell, MA, 1998. Google ScholarDigital Library
- V. Zivojnvic, S. Tjiang, and H. Meyr, "Compiled simulation of programmable DSP architectures," in Proc. IEEE Wkshp. VLSI Signal Processing, May 1995, pp. 73--80. Google ScholarDigital Library
- A. Nohl, G. Braun, O. Schliebusch, R. Leupers, H. Meyr, and A. Hoffmann, "A universal technique for fast and flexible instruction-set architecture simulation," in Proc. ACM/IEEE Design Automation Conf., June 2002, pp. 22--27. Google ScholarDigital Library
- M. Reshadi, N. Bansal, P. Mishra, and N. Dutt, "An efficient retargetable framework for instruction-set simulation," in Proc. IEEE/ACM/IFIP Int. Conf. Hardware/Software Codesign & System Synthesis, Oct. 2003, pp. 13--18. Google ScholarDigital Library
- V. S. P. Rapaka and D. Marculsecu, "Pre-charcterization free, efficient power/performance analysis of embedded and general purpose software applications," in Proc. Design Automation & Test Europe Conf., Mar. 2003, pp. 504--509. Google ScholarDigital Library
- J. Liu, M. Lajolo, and A. Sangiovanni-Vincentelli, "Software timing analysis using HW/SW cosimulation and instruction set simulator," in Proc. Int. Wkshp. Hardware-Software Codesign, Mar. 1998, pp. 65--70. Google ScholarDigital Library
- M. Lajolo, A. Raghunathan, and S. Dey, "Efficient power co-estimation techniques for system-on-chip design," in Proc. Design Automation & Test Europe Conf., Mar. 2000, pp. 27--34. Google ScholarDigital Library
- D. Brooks, V. Tiwari, and M. Martonosi, "Wattch: A framework for architectural-level power analysis and optimization," in Proc. Int. Symp. Computer Architecture, June 2000, pp. 83--94. Google ScholarDigital Library
- W. Ye, N. VijayKrishnan, M. Kandemir, and M. Irwin, "The design and use of SimplePower: A cycle accurate energy estimation tool," in Proc. ACM/IEEE Design Automation Conf., June 2000, pp. 340--345. Google ScholarDigital Library
- G. Qu, N. Kawabe, K. Usami, and M. Potkonjak, "Function-level power estimation methodology for microprocessors," in Proc. ACM/IEEE Design Automation Conf., June 2000, pp. 810--813. Google ScholarDigital Library
- T. K. Tan, A. Raghunathan, G. Lakshminarayana, and N. K. Jha, "High-level energy macro-modeling of embedded software," IEEE Trans. Computer-Aided Design, vol. 21, pp. 1037--1050, Sept. 2002. Google ScholarDigital Library
- C. Brandolese, W. Fornaciari, F. Salice, and D. Sciuto, "Library function timing characterization for source-level analysis," in Proc. Design Automation & Test Europe Conf., Mar. 2003, pp. 1132--1133. Google ScholarDigital Library
- A. Sinha and A. P. Chandrakasan, "JouleTrack - A web based tool for software energy profiling," in Proc. ACM/IEEE Design Automation Conf., June 2001, pp. 220--225. Google ScholarDigital Library
- T. K. Tan, A. Raghunathan, and N. K. Jha, "A simulation framework for energy-consumption analysis of OS-driven embedded applications," IEEE Trans. Computer-Aided Design, vol. 22, pp. 1284--1294, Sept. 2003. Google ScholarDigital Library
- J. Bloomer, Power Programming with RPC. O'Reilly and Associates, Inc., Sebastopol, CA, 1992. Google ScholarDigital Library
- A. V. Aho, R. Sethi, and J. D. Ullman, Compilers: Principles, Techniques and Tools. Addison Wesley Publishing Company, Reading MA, 1986. Google ScholarDigital Library
- J. R. Koza, On the Programming of Computers by Natural Selection.hskip 1em plus 0.5em minus 0.4emrelax The MIT Press, Cambridge MA, 1992. Google ScholarDigital Library
- G. R. Raidl, "A hybrid GP approach for numerically robust symbolic regression," in Proc. Annual Conf. Genetic Programming, July 1998, pp. 323--328.Google Scholar
- P. Long, "Metre v2.3." {Online}. Available: urlhttp://www.lysator.liu.se/c/metre-v2-3.htmlBIBentrySTDinterwordspacingGoogle Scholar
- A. Fraser and T. Weinbrenner, "The Genetic Programming Kernel." {Online}. Available: http://www.cs.ucl.ac.uk/staff/W.Langdon/ftp/weinbenner/gp.htmlGoogle Scholar
- C. Bauer, A. Frink, and R. Freckel, "GiNaC is not a CAS," http://www.ginac.de.Google Scholar
- The GNU Free Software Foundation, "The GNU Scientific Library," http://www.gnu.org/software/gsl/.Google Scholar
- K. Clarkson, "2dch.c." {Online}. Available: http://www.math.niu.edu/ rusin/known-math/96/convhulGoogle Scholar
- B. Chapman and W. Naylor, "wnlib." {Online}. Available: urlhttp://www.willnaylor.com/wnlib.htmlGoogle Scholar
- R. Anderson, "Bipm." {Online}. Available: urlhttp://www.cs.sunysb.edu/ algorith/implement/bipm/distrib/Google Scholar
- W. Qin, "The SimIt-ARM simulator." {Online}. Available: urlhttp://www.ee.princeton.edu/~wqin/armsim.htmGoogle Scholar
- J. Flinn, K. I. Farkas, and J. Anderson, "Power and energy characterization of the ITSY pocket computer (version 1.5)," Compaq Western Research Laboratory, Tech. Rep., Feb. 2000.Google Scholar
- J-L Gailly and M. Adler, "zlib-1.1.14." {Online}. Available: urlhttp://www.gzip.org/zlib/Google Scholar
Index Terms
- Automated energy/performance macromodeling of embedded software
Recommendations
Automated Energy/Performance Macromodeling of Embedded Software
This paper presents an automatic methodology to perform characterization-based high-level software macromodeling. Macromodeling-based estimation can be used to speed up simulation-based software-performance/energy estimation. High-level software ...
Leakage current estimation of CMOS circuit with stack effect
AbstractLeakage current of CMOS circuit increases dramatically with the technology scaling down and has become a critical issue of high performance system. Subthreshold, gate and reverse biased junction band-to-band tunneling (BTBT) leakages are ...
Power macromodeling of global interconnects considering practical repeater insertion
GLSVLSI '04: Proceedings of the 14th ACM Great Lakes symposium on VLSIIn this paper, we present a simple yet highly effective power macromodeling technique for global interconnects that considers optimal repeater insertion. Specifically,our model estimates the interconnect power dissipation from the interconnect length, ...
Reviews
James Edward Tomayko
This paper visits embedded software from two perspectives: reuse and energy consumption. The former is frequently done, and the latter is rarer. The authors describe a tool that macromodels embedded software automatically. Size estimates (revealing complexity) are paired with reuse. The authors make the point that most embedded code is reused, which is good, since its growth is greater than Moore's Law. This tool makes it simpler to both reuse code, and figure out energy consumption and estimates. There is a certain language independence here as well. This paper is best suited for those experienced in this area.
Access critical reviews of Computing literature here
Become a reviewer for Computing Reviews.
Comments