Floating point FP32 (same as IEEE-754 w/ some diffs) core HDL. For Xilinx FPGAs. Include base converters and some math functions. Supported families: Xilinx 6/7 series, Ultrascale, US+. Source files: VHDL
FP WORD 32-bit vector:
EXPONENT - 8-bits. SIGN - 1-bit MANTISSA - 24+1 bits. '1' means hidden one for normalized floating-point values;
Math: A = (-1)^sign(A) * 2^(exp(A)-63) * mant(A)
Component list:
- fp32_fix2float - convert data from INT32 to FP32.
- fp32_float2fix - convert data from FP32 to INT32.
- fp32_addsub - floating point adder.
- fp32_mult - floating point multiplier.