Hi,
I am currently making a tool which will allows me to cache nodes without opening houdini in a custom GUI.There is a difference of speed between the caching if I am using either hython or hbatch. This is what gave me my processor usage when caching with hython:

Whereas if I am using hbatch, this is what I currently have:


Also, this is visible in the time the caching took. For the exact same scene, hbatch takes 2min for 4 frames. For hython, it is taking 4min for 4frames, which is twice the amount of time. The scene on which I am currently experimenting is a High res flip sim.
You can find attached the python script I am using (which is really basic).
Any idea if this is a limitation of python or is it coming directly from Houdini?
By the way, I tested the script with both python2.7 and the beta version of Houdini with python3. And my OS system is Windows

Thanks for the help

I'm also very interested in this since I'm currently investigating my workflow options.
Did you find any solutions / explanations?

Hi Ceonsamea,
Sadly I didn't find any logical explanation as to why it's doing this. The workflow I found out is to start hbatch in a new process from my python script. It doesn't change two much, as the commands are similar