[go: up one dir, main page]
Skip to content

Latest commit

 

History

History
40 lines (30 loc) · 1.08 KB

readme.md

File metadata and controls

40 lines (30 loc) · 1.08 KB
Raw

Floating-Point Parameters - MATLAB Code for Parameters of Floating-Point Arithmetics

About

float_params is a MATLAB function for obtaining the parameters of several floating-point arithmetics. The parameters are built into the code and are not computed at run time.

The parameters are

  • the unit roundoff,
  • the smallest positive (subnormal) floating-point number, xmins,
  • the smallest positive normalized floating-point number, xmin,
  • the largest floating-point number, xmax,
  • the number of binary digits in the significand (including the implicit leading bit),
  • the exponent of xmins,
  • the exponent of xmin,
  • the exponent of xmax

and the arithmetics supported are

  • NVIDIA quarter precision (fp8-e4m3, fp8-e5m2),
  • bfloat16,
  • IEEE half precision (fp16),
  • NVIDIA tf32,
  • IEEE single precision (fp32),
  • IEEE double precision (fp64),
  • IEEE quadruple precision (fp128).

The code was developed in MATLAB R2020a and works with versions at least back to R2016b.

License

See license.txt for licensing information.