Skip to content

Latest commit

 

History

History
36 lines (23 loc) · 1.27 KB

README.md

File metadata and controls

36 lines (23 loc) · 1.27 KB
Raw

FPU - 32-bit single-precision floating point unit

Build Status Build Status

This repo contains code for a simple freeware 32-bit FPU. So far it's just add, subtract, and multiply. It's not even close to being IEEE compliant, but it seems to add 2.0 and 2.0 okay. It's designed to be synthesizable, but I've never done that (yet). For now, it's just RTL that runs correctly on a simulator.

Files include

  • addsub.v - add/sub driver
  • sradd.v - core adder, needs driver to work correctly
  • srsub.v - core subtractor, needs driver to work correctly
  • srmul.v - multiplier, works find standalone

See test dir test/ for examples of how to use the FPU.

For now, see 0notes-fpu.txt for all the gory technicals.

BACKGROUND

What got me here so far:

  • needed a simple 32-bit add/sub/mul unit for my freeware FFT
  • initially I used designware, but that's not free, is it
  • looked for freeware. looked here, looked there. freeware sucked
  • so I wrote my own. it sucks too, but it's mine haha

TODO

See test/TODO.txt