I know the team said they’re working on this, don’t know if it will happen for future updates, but I’m still getting pretty bad performance issues on assemblies and projects (over 2k parts) once steps in history build up. This is on the latest 5.620 version. Collaps history helps some, but same as before, lagging with sketches, selecting and adjusting faces, especially in isolation mode.
Currently developing part of my main project to work with multiple variants of pre-existing parts so lots of downloaded models and lots of modified/created parts going into this. Is only getting worse the more that’s added. Is this just going to be the new norm?
The team said that there were issues with large assemblies and also with lots of feature steps in history adding to the performance, don’t know if this is being addressed. I’m really trying to hop on the HBPM train but this is making it more difficult if every added sketch, step, feature adds in a lot more computing that it can take 5 min to adjust something that used to take less than 1.
Also are there any updates on the large assemblies that had hidden autogenerated backup files? Tried to open the same one I had issues with prior to grab a part out of it and still took 5min to load.
1 Like
Agreed, still experiencing performance issues after update and merging history. It’s a little better but not by much.
In the meantime still using the X_T workaround for large projects.
We’ve fixed most performance issues that were reported, but that doesn’t mean that we can’t make further improvements. Can you be more specific about the issue? Can you share a design confidentially where the performance issue is reproducible with the latest (5.621) version?
That shouldn’t make a difference compared to merging the history. Do you experience performance difference between exporting and importing vs. merging the history?
Interesting, can you share the X_T and the .shapr files? You can send it directly to me or to our support team, we handle your data confidentially.
1 Like
DM’d the Google Drive link. Thanks!
1 Like
Could you make a benchmark design we can download to gauge performance? A stress test per se?
Agreed, I would open up my work but I’ve spent way too much time on it ![:joy: :joy:](https://emoji.discourse-cdn.com/twitter/joy.png?v=12)
I understand also that with large assemblies performance will degrade, but with the introduction of HBPM it’s been at such a rapid pace I can’t find any other reason than that
I know Shapr earlier insisted that retaining all this new information wouldn’t affect performance, now it’s kind of waffley on what type and how much info retained will degrade it. But for anyone that works in large assemblies HBPM is starting to look more and more like a waste.
I for one can’t see the benefit of HBPM applying to a single part when it degrades the performance so much that it is, at this point, a requirement to collapse history. This is just on a legacy model, imported as an x_t to do variants. All the steps of changing geometry to suit new parts is absolutely destroying my iPad and it heats up drastically.