Hello,

Is there any benefit to running a series of nodes in a compiled block. I am not talking about using a compiled block to speed up a for-each loop. Just generally speaking, if a node (or series of nodes) is/are compatible is there any benefit to running them inside a compiled block? Or is it mainly for loops?

thanks

The benefits are mentioned here: https://www.sidefx.com/docs/houdini/model/compile.html [www.sidefx.com]

yep, I read that. I guess I was hoping more for an example of when to use it or for which types of nodes it is helpful/something more specific. As the docs say "The main benefit is multithreaded for-each loops", so I was wondering for something more specific outside of that.

evanrudefx
The main benefit is multithreaded for-each loops", so I was wondering for something more specific outside of that.

The other points right after that

- Another benefit is more efficient use of OpenCL. Normally, even if a node processes geometry on the graphics card, the geometry must be copied back into main memory after each node, because any other node could potentially try to access it. In a compiled block, however, multiple OpenCL-based nodes can keep data on the graphics card as they process it without copying it back, increasing speed.

- In a normal network, each node theoretically makes a copy of the geometry it’s working on. There are lots of optimizations to make this efficient in practice, but it still has a cost. In a compiled block the nodes can work in-place on the same geometry because external references are not allowed. This can give compiled blocks additional speedups

So mostly for data to either stay in GPU memory for chained OpenCL SOPs

And for any other chained SOPs (e.g. chained wrangles) to avoid duplicating their geo in Memory

thanks!

Another use of compiled blocks is for invoke SOPs which should help you simplify networks where the same chain of SOPs might be used in a few places, without having to create an HDA.

thanks, very helpful!