I think Mcneel team has reasons. Be positive.
From my point of view, Rhino cannot handle complex forms a lot of times. What I noticed is that students around me don’t have enough knowledge in Rhinoceros. For instance, when deal with large files, they don’t know how to simplify and manage layers to reduce the file size or even BOX mode. Or, what even worth is that they don’t know how to draw a curve efficiently. Adding unnecessary control points, it certainly is a bad habit of modeling I would say. My friends complain about this a lot of times, but I have no suggestions, because either way (improve modeling skills in short time, or hardware which Rhino cannot take advantage of) seems not imminent. What students end up with is crashing files.
I model things effectively, but I’m not satisfied with the performance Rhino gives. I do think Rhino developers seriously think about the hardware use. In order to have a good projects, most students do want to put money on hardware. So, it doesn’t make sense students are the low profile users. Students care about projects, so they will put money on hardware. Maybe industrial filed is slightly different, but for architecture students, Rhino performance needs to catch up.
I think the Mcneel team knows better than any of us in terms of marketing.
Yes, well thats great information. paraphrasing what I have read regarding Rhino, FlamingoNXT and Bongo: the more GHz on the CPU the better, the more RAM, and the speedier it is the better, and from what you saying about graphics cards, the more Video Memory, the better.
this NURBS system that Rhino uses seems to be focused heavily on using generic parts, So the GTX 780 6GB would actually be the best performer, even though Rhino doesn’t specifically use much of the GPU, the High amount of Video Memory would be beneficial. Am I correct?
It’s also almost $200 cheaper.
The way rhino doesn’t like specialized parts, could be its strength, though currently it seems like a weakness.
O the card that they were reviewing in the article has nothing to do with my interest in it. It is the great big slew of video cards that they test on CAD software. The Firepro W7000 does very well in the OpenGL based application tests you can see.
But I understand now that Rhino doesn’t make a lot of use of OpenGL either. It likes generic parts and specialized driver configurations don’t seem to make much difference.
currently that seems like a weakness, but it may be its strength, as you say.
OpenGL as well as DirectX are both polygon based drawing interfaces. Their conceptual difference is how often 3D geometry needs to be modified and how.
I am not aware of any (pro or gamer) graphics hardware that does natively support NURBS geometry.
So Rhino (and any other NURBS modeler) needs to provide its own conversion and drawing routines for NURBS surfaces and curves. More often than not, this is more efficiently done on the CPU. Part of that conversion is drawing a polygon mesh that approximates the NURBS surface. This is the only part of the drawing process that would really be able to leverage any GPU power.
While Rhino uses the OpenGL interface, it is by no means an OpenGL-App. This “weakness” applies to any software that primarily draws NURBS geometry on generic hardware.
That is a little uncalled for, simply because you can’t decide what video card to get. Just remember you are getting differing advice from users who are most likely all doing different things.
You might want to check out the Holomark thread in which users have benchmarked their cards in Rhino: Holomark 2 Released!
The bottom line is Rhino can make most cards work. How much you want to pay to eke out every last drop is up to you.
I just built a new system around the need to use Rhino. This said here is a reply from the vendor as we were discussing my build.
Another problem I could potentially see is that the video card, the FirePro 2460, is not a very powerful video card and since you would primarily be using this system for 3D design,
you might want a more powerful video card. The Asus GeForce GT 640 Passive would be an option that, while it only has support for three monitors instead of 4 that the FirePro has, has more video memory (2 GB) than the FirePro does.
A company called Puget systems built this new system and they worked with me prior to the build. I had an old graphics machine and was sorely in need of a new machine. I built his system around a an I7 processor and with a solid state boot drive I am thrilled with the performance of this new system. Go online and shop a machine at Puget, it cost no money until you say yes and the staff there goes out of their way to make sure you get exactly what your after.
My renders from the Rhino Render- on a laptop that has 3.5" floppy drive installed on it mind you- have sold literally millions of dollars worth of projects, with 10-20 min sprucing up on the primary in PS, and nothing at all done to the supporting images. We are all busy people after all, selling stuff, right? Who the hell wants to blow extra time on proposals?
In your case especially, I can promise you Rhino or it’s rendering performance is not the bottleneck.
First of all… more than 4Gigs of VRAM available is… nice.
My question was really aimed at the intended usage. There may be situations where more RAM is of help. From my own experience it’s nice but not neccesary. Depending on the machine I’m using, I have 1 or 2 gigs at my disposal and while I have seen some heavy graphics slowdown, memory consumption was never the problem.
Then again, I’m not sure about the very latest release, but Photoshop shouldn’t use very much of GPU memory.
I think you’re getting dangerously close to losing your rights to post in here…