Skip to content

JeffreySarnoff/XFloats.jl

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

186 Commits

.github/workflows

.github/workflows

 
 

benchmarks

benchmarks

 
 

src

src

 
 

test

test

 
 

LICENSE

LICENSE

 
 

Project.toml

Project.toml

 
 

README.md

README.md

 
 

Repository files navigation

XFloats.jl

Copyright © 2017-2020 by Jeffrey Sarnoff. This work is released under The MIT License.

XFloats exports two these two types: XFloat16, XFloat32.

  • These floating point types correspond to Float16 and Float32.
  • Use them to compute results with greater accuracy at very high speed.

benchmarks

           relative speeds
           X32∝F32  X64∝F64

    add:     1.00    1.08
    mul:     1.08    1.08
    fma:     0.93    1.08
other extended precison floating point types
  • The more accurate type for Float64 is Double64 (DoubleFloats.jl).
    • It calculates more slowly than these types do.

About

Precision-doubled floating point types nearly as performant as hardware floats.

Topics

performance julia accuracy floating-point

Resources

Readme

License

MIT license
Activity

Stars

2 stars

Watchers

2 watching

Forks

1 fork
Report repository

Releases

2 tags

Packages

No packages published

Languages