Floating Point Unit Implementation on FPGA

Free registration required

Executive Summary

As densities of FPGA are increasing day by day, the feasibility of doing floating point calculations on FPGAs has improved. Moreover, recent works in FPGA architecture have changed the design tradeoff space by providing new fixed circuit functions which may be employed in floating-point computations. By using high density multiplier blocks and shift registers efficient computational unit can be developed. This paper evaluates the use of such blocks for the design of floating-point units including adder, subtractor, multiplier and divider.

  • Format: PDF
  • Size: 621.5 KB