I'm following this project for months, I had a few insignificant contributions and some test about performance; syntax..., what looks promising is the philosophy of "one way to do things", which allows to produce a smaller yet faster compiler and a predictable code and increase compiler's performance aswell as overall future tools (e.g. frameworks, libraries, translation tools C/C++).
I believe that many innovations has been made by removing things and not by adding them.
I can't wait to see how this gonna be performing on heavy machine learning frameworks (tensorflow, pytorch), maybe I'll do some matrix manipulations as soon as v1.0 becomes available to see how things is gonna work.
Sorry! I don't have any serious performance tests to share, just playing around with the syntax gist.github.com/bitsnaps/98b59cf8f...
most of the tests has been made in the project at compiler/tests/
I'm following this project for months, I had a few insignificant contributions and some test about performance; syntax..., what looks promising is the philosophy of "one way to do things", which allows to produce a smaller yet faster compiler and a predictable code and increase compiler's performance aswell as overall future tools (e.g. frameworks, libraries, translation tools C/C++).
I believe that many innovations has been made by removing things and not by adding them.
I can't wait to see how this gonna be performing on heavy machine learning frameworks (tensorflow, pytorch), maybe I'll do some matrix manipulations as soon as v1.0 becomes available to see how things is gonna work.
I would like to look at your performance tests. Indeed removing looks like more interesting part, than adding.
Thanks for commenting!
Sorry! I don't have any serious performance tests to share, just playing around with the syntax
gist.github.com/bitsnaps/98b59cf8f...
most of the tests has been made in the project at
compiler/tests/