/float754

IEEE 754-2008 floating point emulation and hardware

Primary LanguageCOtherNOASSERTION

IEEE 754-2008 Floating point emulation and hardware

This project has two objectives

  1. Fast C code for emulating IEEE 754-2008 floating point operations
  2. Based on 1. an efficient floating point hardware in VHDL

Roadmap

  1. C implementation of half precision: add, multiply, divide
  2. Complete brute force checking of half precision arithmetics (2^32 cases)
  3. C implementation of single precision
  4. C implementation of double precision
  5. C implementation of quad precision
  6. Rearrangement of C code to better fit hardware demands
  7. Translation to VHDL

Open Questions

  • Should exceptions be supported?
  • More than one rounding mode?
  • Decimal support?

License

Licensed under the ISC licence (similar to the MIT/Expat license).