Compact Numerical Function Generators Based on Quadratic Approximation: Architecture and Synthesis Method

This paper presents an architecture and a synthesis method for compact numerical function generators (NFGs) for trigonometric, logarithmic, square root, reciprocal, and combinations of these functions. Our NFG partitions a given domain of the function into non-uniform segments using an LUT cascade, and approximates the given function by a quadratic polynomial for each segment. Thus, we can implement fast and compact NFGs for a wide range of functions. Experimental results show that: 1) our NFGs require, on average, only 4% of the memory needed by NFGs based on the linear approximation with non-uniform segmentation; 2) our NFG for 2x-1 requires only 22% of the memory needed by the NFG based on a 5th-order approximation with uniform segmentation; and 3) our NFGs achieve about 70% of the throughput of the existing table-based NFGs using only a few percent of the memory. Thus, our NFGs can be implemented with more compact FPGAs than needed for the existing NFGs. Our automatic synthesis system generates such compact NFGs quickly.

[1]  Shinobu Nagayama,et al.  Compact representations of logic functions using heterogeneous MDDs , 2003, 33rd International Symposium on Multiple-Valued Logic, 2003. Proceedings..

[2]  Florent de Dinechin,et al.  Second Order Function Approximation Using a Single Multiplication on FPGAs , 2004, FPL.

[3]  Shinobu Nagayama,et al.  Programmable numerical function generators: architectures and synthesis method , 2005, International Conference on Field Programmable Logic and Applications, 2005..

[4]  Jack E. Volder.  The CORDIC Trigonometric Computing Technique , 1959, IRE Trans. Electron. Comput..

[5]  Tsutomu Sasao.  A cascade realization of multiple-output function for reconfigurable hardware , 2001 .

[6]  John H Mathews,et al.  Numerical methods for computer science, engineering, and mathematics , 1986 .

[7]  Shinobu Nagayama,et al.  On the optimization of heterogeneous MDDs , 2005, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[8]  Jie Cheng,et al.  High-performance architectures for elementary function generation , 2001, Proceedings 15th IEEE Symposium on Computer Arithmetic. ARITH-15 2001.

[9]  Nagayama Shinobu,et al.  Programmable Numerical Function Generators Based on Quadratic Approximation: Architecture and Synthesis Method , 2006 .

[10]  Michael J. Schulte,et al.  Approximating Elementary Functions with Symmetric Bipartite Tables , 1999, IEEE Trans. Computers.

[11]  J.-M. Muller,et al.  A new scheme for table-based evaluation of functions , 2002, Conference Record of the Thirty-Sixth Asilomar Conference on Signals, Systems and Computers, 2002..

[12]  Michael J. Schulte,et al.  Symmetric bipartite tables for accurate function approximation , 1997, Proceedings 13th IEEE Sympsoium on Computer Arithmetic.

[13]  Wayne Luk,et al.  A Gaussian noise generator for hardware-based simulations , 2004, IEEE Transactions on Computers.

[14]  Tsutomu Sasao,et al.  Realization of multiple-output functions by reconfigurable cascades , 2001, Proceedings 2001 IEEE International Conference on Computer Design: VLSI in Computers and Processors. ICCD 2001.

[15]  Ray Andraka,et al.  A survey of CORDIC algorithms for FPGA based computers , 1998, FPGA '98.

[16]  Jean-Michel Muller,et al.  Elementary functions - algorithms and implementation (2. ed.) , 2006 .

[17]  Tsutomu Sasao,et al.  Application of LUT cascades to numerical function generators , 2004 .

[18]  Michael J. Schulte,et al.  The Symmetric Table Addition Method for Accurate Function Approximation , 1999, J. VLSI Signal Process..

[19]  Wayne Luk,et al.  Non-uniform Segmentation for Hardware Function Evaluation , 2003, FPL.

[20]  Takashi Horiyama,et al.  Minimization of fractional wordlength on fixed-point conversion for high-level synthesis , 2004, ASP-DAC 2004: Asia and South Pacific Design Automation Conference 2004 (IEEE Cat. No.04EX753).

[21]  Takashi Horiyama,et al.  Minimization of fractional wordlength on fixed-point conversion for high-level synthesis , 2004 .

[22]  Naofumi Takagi,et al.  Function evaluation by table look-up and addition , 1995, Proceedings of the 12th Symposium on Computer Arithmetic.