I have a very large model and I’m exploding the whole thing (I know its kind of a crazy thing to do)
but my PC is showing my hardware use as very low. Sketchup is still processing this task (or maybe its crashed? Hard to tell) but should it be using more computing power is there a way to get it to use more to speed up the task?
Sketchup, and for that mater all 3d applications, only use one core of your processor. Which is why for years it has been said a high single core speed is better than multiple cores of lower speeds.
It’s the nature of the best. May change with quantum computing but until then it is bottlenecked by the single core. It does however mean you can do other things, even other instances of sketchup, simultaneously on the other cores. Something I often do, if I know something is going to choke I have a separate file to work on while it chugs away.
Thanks Box. Yes I have read similar things in the past I guess I’m just seeing the effects personally. I’ll leave it chugging away in the background and keep my fingers crossed it doesn’t crash.
It’s sort of one of the reasons why components can be so useful. You can have a huge site plan and break it into various components. You can then work on those components and simply reload them into the main model when needed. That could even be a house and each room is a component and only used in relevant scenes.
Many many ways to optimise.
If you click on the CPU you can see the load on each core - SketchUp will only be operating on one ofthose, so 5 or 6 % of the total for a 24 core CPU is likely one core working solidly and the some other bits from other processes.
This is a very simplified (and possibly wrong!) way to think about it
And with that logic, could it be possible for SketchUp to undertake one operation within a component, and while other operations are taking place in differernt components?
Imagine this - you run an operation (let’s say it’s Explode) and the component you run it on becomes temporarily locked. Once the operation is complete that component becomes unlocked. The geometry/data would have to then be merged with the rest of the model hierarchy.
I don’t know if this would even save much time at all! Most SKP operations are pretty quick these days. Interesting to think if it’s possible, though.