The Integer to Floating Point Unit core performs integer to IEEE 754, single-precision, floating-point conversion. The operand is a 32-bit integer, which is either signed (2s complement format) or unsigned (binary format). The result is 32-bits wide and consists of 1 sign bit, 8 exponent bits and 23 fraction bits.
The exponent is biased, with a bias of 127. The leading bit of the fraction is implicit for normal numbers.
- Conversion of integers to IEEE-754 single-precision floating point values.
- Support for signed and unsigned integers.
- Status flag indicating inexact operations.
- Simple interfacing
- Low gate-count
- High throuhput
- Full source code
- Test harness
- Synthesis scripts for Xilinx, Altera and ASIC
- High performance computing
- Beam forming
- Wireless - MIMO/SDR
- Energy efficiency computing