E3SM-Project / EKAT

Tools and libraries for writing Kokkos-enabled HPC C++ in E3SM ecosystem
Other
15 stars 7 forks source link

Consider adding support for half-precision floating point numbers. #9

Open jeff-cohere opened 4 years ago

jeff-cohere commented 4 years ago

Is your feature request related to a problem? Please describe. Some projects have been toying with the idea of using half-precision numbers, in addition to single- and double-precision ones. It might be nice to enable EKAT to work with these.

Describe the solution you'd like To let EKAT work with half-precision numbers, we'd generalize some of the logic in ekat_type_traits.hpp near line 29. Instead of using a double to indicate that double precision should be used, we'd use an enum to select the half, single, or double precision. See this conversation.

welcome[bot] commented 4 years ago

Thanks for opening your first issue here! Be sure to follow the issue template!