search

eagles-project / haero

A toolbox for constructing performance portable aerosol packages
Other
3 stars 3 forks source link

Evaluate using "KPP" as an alternative for chemical kinetics #82

Closed jeff-cohere closed 3 years ago

jeff-cohere commented 3 years ago

Po-Lun mentioned that we should take a look at the Kinetic PreProcessor, which is what MAM has been using for atmospheric chemistry. It seems like KPP has been reimagined as FlexChem. And then again, some folks are using their own source-to-source translator to convert KPP-generated files to CUDA source files for running on GPUs.

All the stuff I can find about KPP is pretty confusing. An "official looking" repo for KPP is here:

https://github.com/geoschem/KPP

However, it doesn't look very active. Maybe we can ask Po-Lun and Kai for people to consult about what's being done in the KPP world at our next meeting.

An interesting note: According to this file and this file, KPP is licensed under the GPL, meaning all non-released software that uses it is in violation of the license. It wouldn't surprise me to learn that this is true. This is something we should mention to Po-Lun.

jeff-cohere commented 3 years ago

The license for KPP prevents us from using it within our model. We'd have to incorporate generated files. Maybe we could surface an option to make this practical, but there'd be no way to distribute KPP or use it as anything but part of a chemistry workflow.