Hey,

I've been following through the Spiro Toy tutorial. (https://www.sidefx.com/index.php?option=com_content&task=view&id=2961&Itemid=410) [sidefx.com]
I'm curious as to why two ways of creating the same setup create very different calculation speeds.

When I build my setup with the popnet and Connect Adjacent Pieces nodes in the same geometry node it runs very slowly. (Spiro 3.hiplc)

When I follow how Ari does the setup in the tutorial (separate nodes on scene level for each sops setup and dops) it runs much faster. (Spiro 3 fast.hiplc)

In fact the Connect Adjacent pieces doesn't even really function in the first setup.

Is it best to not create POP Nets inside Geo and keep them in their own setup? Or is it the issue of the Connect Adjacent which is the issue. Really I'm just interested in best practice and understanding how it works.

Thanks

in terms of speed it doesn't matter that much whether you use separate networks or the same one

the difference in your files is that in the slow one you use 4 popnet substeps and you source 5000 particles/s (most of them overlapping) which is much more than the fast one which sources just 24/s
then obviously connect adjacent pieces took much longer time and without visible results as most particles that it connected were at the same position

tamte
in terms of speed it doesn't matter that much whether you use separate networks or the same one

the difference in your files is that in the slow one you use 4 popnet substeps and you source 5000 particles/s (most of them overlapping) which is much more than the fast one which sources just 24/s
then obviously connect adjacent pieces took much longer time and without visible results as most particles that it connected were at the same position

Aaah thanks very much.