I'd like to know how performance really be enhanced in SolidWorks when the real problem is that the software components used to create SolidWorks don't properly support multi-core processing?
I've written about this subject extensively on my blog and I fail to comprehend how any SolidWorks user can expect SolidWorks performance to really be enhanced until concurrency in the constraint solvers and the kernel that SolidWorks users is fully supported.
SolidWorks V6 will use the CGM kernel which does support concurency. I've asked Dassualt on Twitter for proof of performanc in Cata which uses the CGM kernel and I have not gotten any response. Neither have others who has asked for proof of CGM kernel performance in Catia.
At 25:54 in the link below a claimed major advantage for the CGM kernel is being able to roll back and forth in the history tree very quickly. From many years of using Solidworks, I know that this is a major problem even on modest sized part models. I would be interested in knowing if anyone has done a direct comparison between Catia V5 / V6 (CGM kernel) and Siemens NX (Parasolid kernel) or SolidWorks (Parasolid) in this area.
I think it's very wrong that Dassualt continues to ignore the root of what I consider to be the problem and refuses to back up claims that its Spatial divsion is making about the CGM kernel. At this point I expect answers where none have been forthcoming.