The folks at a new start-up called OptNgn say they are offering a floating-point VHDL library under the GPLv3 Open Source License. FPGA designers can now save months of coding and debug time by using these floating-point functions instead of creating their own VHDL from scratch.
The parameterized floating point operators being introduced are in three groups: Arithmetic, Transcendental, and Trigonometric, as illustrated below:
Also included with the VHDL source code are test-benches, ModelSim command files, and Xilinx synthesis scripts to aid in verification and integration to an FPGA project. Users can expect these operators to run at 200 MHz when using the Xilinx Virtex-5 series.
"I see tremendous synergy between the OptNgn libraries and our C-to-FPGA compiler," said David Pellerin, CTO of Impulse Accelerated Technologies. "The combination of hardware compiler and optimized libraries enables hardware accelerated embedded systems for image processing, DSP and general purpose FPGA acceleration."
This library is a sanctioned modification and enhancement of the respected and vetted FPLibrary developed by the Arénaire project, at ENS Lyon. To download the latest version of the floating-point library visit Source Forge at: http://libhdlfltp.sourceforge.net.