3D Graphics Performance Comparison – Part 11 (IRONCAD)

<< Part 10

I think I can say that this 3D graphics comparison series is beginning to have a real effect on some CAD vendors. I know of at least one CAD vendor who has already acted upon the issue I raised about surprisingly high CPU utilization of some CAD systems when moving the mouse in an empty graphics window. This issue was actually brought to my notice by a reader and resulted in a blog post titled “CPU Usage In Idle CAD Systems“.

My initial comparison showed that the default graphics settings of some CAD systems were not actually optimal. Also as I continue to add parts to this series, some CAD vendors are actually tweaking their graphics engines. I know that because they are sending me beta versions of their software to test my engine model with and I am giving them feedback.

One such CAD vendor is IronCAD. I have been working closely with VP of Marketing Cary O’Conor for a few days now. To give you an idea of how close, last night after I returned from my weekly Friday night boozing session with friends, Cary did a web meeting with me to show the latest stuff that his guys had come up with. I was quite a few pegs down and the clock in my house was trying its best to tell me that it was 2:30 am. But once I started seeing the stuff on Cary’s computer the effect of the alcohol evaporated and my eyes almost popped out. Here’s why.

Two words. Freaking Awesome! This is a beta of IRONCAD 2011 using the default graphics settings. Now compare this with the video from my initial comparison of IRONCAD 2009 using its default graphics settings.

Poles apart. Just poles apart. In one of our discussions Cary and I spoke about SpaceClaim’s nice little feature of automatically coloring parts when the STEP file does not contain color information. And look, within a couple of days IRONCAD is doing the same thing. The reason parts were dropping off in the original video was because IRONCAD’s default graphics settings were not optimized for large models. Now they seem to be. Also IRONCAD had a weird option of delaying the display of edges after the user released the mouse button to stop navigating. I never really understood why someone would want to delay the display of edges. It gives the impression that the software is busy calculating the edges, when actually it is just sitting idle waiting to get the green signal to start calculating edges. In IRONCAD 2011 this time delay is set to 0 by default and edges are displayed immediately after the user stops navigation.

Here is a screen shot that shows the kind of display quality that you get with the default graphics settings for IRONCAD 2011. Pretty good, I’d say.


Click image for larger view

Today I got an email from someone asking me what was the point of doing this graphics performance comparison. He wrote, “I suspect that, for most users (with reasonably good systems) it’s not going to be a significant enough differentiator to cause them to consider changing CAD software.

I replied, “I think graphics performance is one of the last reasons why someone would want to switch CAD systems. They will be much better off upgrading their hardware instead. I really don’t have any motives behind this comparison. I’m simply laying things out there for all to see and draw their own conclusions if they want to or agree/disagree with mine. We all know how CAD vendors make outrageous claims about 10x faster performance in this area and that area by twisting the results of some “independent” paid study to support those claims. Sometimes it is nice to hand them all a mirror and let them see where they really stand with respect to others.”

And of course, if this kind of a comparison encourages CAD vendors like IronCAD to actually go ahead and improve their software, then nothing like it.

Part 12 >>

  • http://twitter.com/CADkid Marijn

    So IronCAd optmized there software for small assemblies? That makes absolute no sense. f it is optimized for big assemblies wouldn't it just also be optimal for small assemblies?

  • http://www.deelip.com Deelip Menezes

    Yes, graphics is one of those things where you need to be ready for the bad cases. The good cases will take care of themselves automatically.

  • gale

    Deelip you rock. After you post whole story, the CAD companies will pump their heads and it will be new age of CAD graphics optimization.

  • http://www.deelip.com Deelip Menezes

    LOL! I don’t think this series will have that kind of “catastrophic” consequences. ;-) But if it makes even a slight difference, then it’s a good thing.

  • Cary O'Connor

    Hi Marijn,

    The optimization in IRONCAD 2011 was for the core graphics for large assemblies. What you are seeing is how it affects all the way down to this smaller assembly. This test model is only one of many that we are using for optimization. We have much larger data sets that we have in our stable. Also, we are tweaking other area that affect the overall experience and not just the graphics side. The 2011 release should easily feel like a big step in overall performance.

    Cary

  • http://twitter.com/dailycadcam DailyCADCAM

    Thanks for such a good post, one of best series of articles I have read. I think this posts will definitely have some impact on improvement of CAD graphics.

  • http://www.deelip.com Deelip Menezes

    Glad you liked it.

  • Renu Srivastava

    NICE DESIGN RENU




Archives

© 2014 Deelip.com. All Rights Reserved. Deelip.com is a registered trademark of Deelip Menezes. Log in - Designed by Gabfire Themes