ABSTRACT
We consider synthesis of arithmetic DSP circuits with finite precision fixed-point operations. The aim is to choose the lowest cost implementation that matches a real-valued specification within the allowed imprecision. Starting from Taylor Series or real-valued polynomials, we demonstrate first a method to obtain satisfying implementations that uses intermediate Arithmetic Transform polynomials as an analytical apparatus suitable to precision analysis for both the quantization (bit-width) and approximation sources of imprecision. We then derive the precision optimization algorithm that explores multiple precision parameters in a branch-and-bound search.
- A. Gaffar, O. Mencer, W. Luk, and P. Cheung, "Unifying bit-width optimisation for fixed-point and floating-point designs," in Proc. IEEE Symp. Field-Programmable Custom Comput. Mach, FCCM 2004, pp. 79--88. Google ScholarDigital Library
- S. Kim and W. Sung, "Fixed-point error analysis and word length optimization of 8 x 8 IDCT," IEEE Trans. Circuits Syst. Video Tech., Vol. 8, No. 8, Dec. 1998, pp. 935--940. Google ScholarDigital Library
- K. Kum and W. Sung, "Combined word-length optimization and highlevel synthesis of digital signal processing systems," IEEE Trans. CAD Vol. 20, No. 8, Aug. 2001, pp. 921--930. Google ScholarDigital Library
- C. Shi and R. Brodersen, "Automated fixed-point data-type optimization tool for signal processing and communication systems," Proc. Design Automation Conf. 2004, pp. 478--483. Google ScholarDigital Library
- M. Willems, V. Bürgens, H. Keding, T. Grötker and H. Meyr, "System level fixed-point design based on an interpolative approach," Proc. Design Autom. Conf. 1997, pp. 293--298. Google ScholarDigital Library
- Authors, "Paper Title Withheld for blind review", Published in Journal of (name of publication withheld for blind review).Google Scholar
- A. Nayak, M. Haldar, A. Choudhary, P. Banerjee, "Precision and error analysis of Matlab applications during automated synthesis for FPGAs," Proc. DATE, 2001, pp. 722--728. Google ScholarDigital Library
- G. Constantinides, P. Cheung, and W. Luk, "Wordlength optimization for linear digital signal processing," IEEE Trans. on CAD vol. 22, no. 10, pp. 1432--1442, Oct. 2003. Google ScholarDigital Library
- D.-U. Lee, A. Gaffar, R. C. C. Cheung, O. Mencer, W. Luk, and G. Constantinides, "Accuracy-Guaranteed Bit-Width Optimization", IEEE Trans. CAD, Vol. 25, No. 10, Oct. 2006, pp. 1990--2000. Google ScholarDigital Library
- Smith, J.; De Micheli, G.; "Polynomial Circuit Models for Component Matching in High-level Synthesis", IEEE Trans. VLSI, Vol. 9, No. 6, Dec. 2001. pp. 783--800. Google ScholarDigital Library
- Pang, Yu; Radecka, Katarzyna; Zilic, Zeljko; "Arithmetic Transforms of Imprecise Datapaths by Taylor Series Conversion", Electronics, Circuits and Systems, 2006. ICECS '06. 13th IEEE International Conference on 10--13 Dec. 2006 Page(s):696--699Google Scholar
- L. de Figueiredo and J. Stolfi, "Self-validated numerical methods and applications," in Brazilian Mathematics Colloquium Monograph. Rio de Janeiro, Brazil: IMPA, 1997.Google Scholar
- S. Wadekar and A. Parker, "Accuracy sensitive word-length selection for algorithm optimization," in Proc. IEEE Int. Conf. Comput. Des., 1998, pp. 54--61. Google ScholarDigital Library
- R. Cmar, L. Rijnders, P. Schaumont, S. Vernalde, and I. Bolsens, "A methodology and design environment for DSP ASIC fixed point refinement," in Proc. ACM/IEEE Design Automation Test in Europe. DATE '99, 1999, pp. 271--276. Google ScholarDigital Library
Index Terms
- Optimizing imprecise fixed-point arithmetic circuits specified by Taylor Series through arithmetic transform
Recommendations
High-Level Design of Multiple-Valued Arithmetic Circuits Based on Arithmetic Description Language
ISMVL '08: Proceedings of the 38th International Symposium on Multiple Valued LogicThis paper proposes a high-level design method of multiple-valued arithmetic circuits. The proposed method uses a cell-based approach with a dedicated hardware description language called ARITH. By using ARITH, we can describe and verify any binary/...
Optimization of imprecise circuits represented by Taylor series and real-valued polynomials
Arithmetic circuits in general do not match specifications exactly, leading to different implementations within allowed imprecision. We present a technique to search for the least expensive fixed-point implementations for a given error bound. The method ...
Comments