As the title said. Why not the Blender developing team take a look and see how it can improve the Blender’s python code.
1 Like
I think before we take a look at how fast the Python code runs, it is probably better to ask how performance critical is this Python code?
For example, if you have a code that you run maybe once a week and takes at most a second to evaluate, then it may not be worth investigating the performance. However, if the code runs all the time, or needs to be as close to realtime as possible, then it is really worth investigating.
The tool in the link might be interesting to investigate, though. I’d be wary of anything that says that “Project X is the silver bullet for all your needs”, since very few things are really game changing these days. Only way to find out is to do some research and test this out for yourself.
2 Likes
The funny thing about profilers is that they are… quite slow 
Profilers are a huge drain of resources, analyzing code is no simple task, and in many cases, the time necessary to use a profiler far outweighs run or development time.
Also, I’m curious- what leads you to assume the developers don’t already use profilers? I can assure you, all the developers I personally know do 
2 Likes
An AI profiler is a cool idea, but I’m not sure it helps in this case.
It would be nice to see Julia added to Blender instead of focusing on Python.
Have you ever looked into the sources… or how blender is build ?
Blender is written in C/C++ and the python code is used for the gui and to make addons…
1 Like