A flexible floating-point format for optimizing data-paths and operators in FPGA based DSPs

ABSTRACT

Video signal processing requires complex algorithms performing many basic operations on a video stream. To perform these calculations in real-time in a FPGA, we must use innovative structures to meet speed requirements while managing complexity. As part of a project aiming at the development of a video noise reducer, we developed an optimized processing stream that required some floating-point calculations. This paper presents the rationale for developing a floating-point unit, justifies the data representation used, its implementation in a Xilinx VirtexE FPGA and reports the performance we obtained. A divider using this representation is also presented, with its implementation and performances in the same FPGA.

REFERENCES

• 1: ANSI/IEEE Std 754-1985 "IEEE Standard for Binary Floating-Point Arithmetic", 1985.
• 2: Connors, D.A., Yamada, Y. and Hwu, W-M.W. "A Software- Oriented Floating-Point Format for Enhancing Automotive Control Systems", August 1999.
• 3: Kelley, M.J. and Postiff, M.A. "A CMOS Floating-point Unit" 17th Annual Student VLSI Design Contest, Experienced Class Entry, 1997.
• 4: Ligon III, W.B., McMillan, S., Monn, G., Stivers, F and Underwood, K.D. "A Re-evaluation of the Practicality of Floating-Point Operations on FPGAs".
• 5: Stuart F. Oberman , Michael J. Flynn, Division Algorithms and Implementations, IEEE Transactions on Computers, v.46 n.8, p.833-854, August 1997 [doi>10.1109/12.609274]
• 6: POWER2 Floating-point Unit: Architecture and Implementation, http://www.austin.ibm.com/tech/fpu.html.
• 7: Sahim, I., Gloster, C. and Doss, C. "Feasibility of Floating- Point Arithmetic In Reconfigurable Computing Systems" MAPLD on Adaptative Computing, September 2000, vol.3.
• 8: N. Shirazi , A. Walters , P. Athanas, Quantitative analysis of floating point arithmetic on FPGA based custom computing machines, Proceedings of the IEEE Symposium on FPGA's for Custom Computing Machines, p.155, April 19-21, 1995
• 9: Iakovos Stamoulis , Martin White , Paul F. Lister, Pipelined Floating Point Arithmetic Optimized for FPGA Architectures, Proceedings of the 9th International Workshop on Field-Programmable Logic and Applications, p.365-370, August 30-September 01, 1999
• 10: The SPARC Architecture, http://www.cs.earlham.edu/ mutioke/cs63/sparc.htm#4.

INDEX TERMS

Keywords:
FPGA, data-path optimization, floating-point, floating-point/fixed-point conversion, hardware division, hyardware optimization, video-processing