Rebuild time benchmarks?

S1mon

We're trialing Onshape against Solidworks. Are there any public benchmarks of rebuild times for similarly modeled parts in Solidworks and Onshape?

alnis

I've attached a SolidWorks version of a model I did for practice that I also modeled in almost the same way in Onshape. When making a trivial edit to one of the dimensions at the top of the tree to trigger a rebuild without usable cached data, the rebuild time was 10 seconds in SolidWorks and 15 seconds in Onshape (timed with a stopwatch from accepting the sketch to having the model onscreen).
Link to Onshape version (the configurable one is still WIP and does not work correctly):
https://cad.onshape.com/documents/49884d6ffce0385868d6e84a/v/124d708ec1d9f2de66669212/e/e231f1f7cf16c66c1093751b

While SolidWorks does appear faster in this case, here are some performance considerations:

1. SolidWorks crashed a few times while modeling this part, causing me to lose about an hour of work because my last save corrupted.
2. Consider the time to open and get to work on these files. (How long does it take for the Onshape link to open and to make a copy of the document? How long does it take to download the file, add it to your PDM system, open SolidWorks, and open the part?)
3. In Onshape, manual rebuilds are never necessary. The system ensures that the model onscreen is always built with the latest user inputs, even with complex dependency chains and large assemblies. Therefore, rebuild time is only really relevant when editing the top features in a feature tree.

Also, it is my understanding that Onshape parallelizes the rebuilding of large assemblies by rebuilding each part studio independently (e.g. after changing a configuration which affects all/many parts in the assembly), while SolidWorks must do all of this on one thread (thus hindering performance on large projects), but I am not 100% sure on this.

Of course, you can see that I do have a preference for one of these two programs, but hopefully, this helps with judging their performance

lana

@alnis_smidchens
Thank you! That is a very useful benchmark.