Are OpenGL version conflicts a problem


#1

I am thinking about getting the Firepro W7000, and I absolutely must know if it is worth it when using Rhino 5.


#2

Version conflicts generally are a problem. That’s why they are called “conflicts”.

Seriously, is there any specific conflict situation, you are referring to?


#3

I am referring to performance relationships between hardware and software.

I have heard reports that this card will comfortably allow users to push 30,000,000 polygons. Its an awesome card for those who use OpenGL applications say the reports. Does this hold true with Rhino as well?

by version conflicts are you telling me that Rhino 5 is really only designed to work optimally with 10 year old video cards?


#4

OpenGL isn’t really meant to draw NURBs Objects. While there are some interfaces that expose NURBs functionality, they basically create a render mesh and draw that the default way. Since Rhino manages its own render meshes, that’s not depending on OpenGL here.

The Rhino display pipeline does a great deal more than just pushing polygons, like drawing curves. On the other hand, that 10 year old part of the OpenGL interface didn’t change at all. Newer versions may include more fancy shader and texturing stuff, but Rhino just doesn’t use that.

Long answer short: if you are modelling in rendered mode, you may be able to see some of the OpenGL speed regardless of OpenGL version. If not, the bottleneck will not be the graphic card.

From personal experience, I haven’t seen Rhino make good use of any current GPU and while I only use consumer grade cards tailored to DX, I doubt that this is due to bad OpenGL support.


#5

Possibly I don’t even know how to ask this question correctly, because I do not understand API’s very at all, I am really judging by performance. I just look at the bottom line.

is Rhino limited to about 5,000,000 polygons even though one may own a card that is capable of pushing over 30,000,000 polygons? And this has something or other to do with the old version of OpenGL it uses?


#6

The short answer is no, Rhino shoves however many polygons at the video card as it has and it cranks them out as fast as it can. Your number of “pushing 30 million polygons” is fairly meaningless anyway… how may per second? In what app? With what sort of lighting?

I wouldn’t frankly recommend a ludicrously expensive video card just for Rhino, and it’s not raw polygon-pushing for which you buy such a beast vs a “gaming” card anyway, it’s for accelerating other graphic features like wireframes.


#7

q1) How many per second?: we’ll in real development the polygons would likely spike very sharply at one point as user uses various array tools to blast out shapes and patterns, then lots of fiddling kind of work were a smooth display port experience is needed.

q2) In what app?: I don’t fully understand this question your asking, isn’t rhino the app? Rhino is certainly not the only program I am using. i am using FlamingoNXT, Bongo(I don’t have the 2.0 version yet, likely get it soon though). And I am studying Blender(which fully utilizes OpenGL), GIMP, and some other things as well.

As far as my limited knowledge apprehends. Rhino is the main technical drawing center of the whole design flow process. The rhino environment is also where a lot of other studies can be accomplished. I expect my drawings to be very detailed, maybe more detailed than some firms would consider necessary, and this is because I have less infrastructure.

q3) with what sort of lighting?: the lighting I am thinking about is just the default working environment lighting.

I wouldn’t frankly recommend a ludicrously expensive video card just for
Rhino, and it’s not raw polygon-pushing for which you buy such a beast
vs a “gaming” card anyway, it’s for accelerating other graphic features
like wireframes.

I don’t know what your talking about here. Is there something wrong with the Rhino software?

In video card tests the w7000 was 3.5x better than the GTX cards on CAD programs using OpenGL. GTX cards got a pathetic 25 FPS on the viewport test, while the w7000 got around 85 FPS. So it seems stupid to go with the GTX cards. Unless there is something I don’t know. The test was only using a measly 1000 cubes at a time to test the cards. So I seriously do not know what you are talking about.

Also I may not fully know what the right questions to ask are. I have used File sizes and polygons counts as a way of roughly gathering performance data. you probably know of a better things to Judge the performance of software with?

I really really need to understand the box that I am working in, the limitations of the environment I am working in. I start with the limits and then work backwards with what i am doing.

I really don’t understand why you think that’s a lot. But these software products are really just sales tools for conveying information, idea, weeding out problems. So my work needs to look good, meaning more polygons, more everything.

Its NOT how many polygon do I need to build the thing. It is how many polygons do I need to SELL the thing.


#8

What I meant was, wherever you got that “it can handle 30million polygons” number, that’s in a certain context of some application(that probably wasn’t Rhino) under some specific conditions at a certain number of frames per second. We have no context, it’s a useless number.

CAD benchmarks are specifically useless for Rhino since many other such programs are designed around supporting “pro” cards specifically and are going to be specifically bad on anything else. Rhino’s not like that, it’s designed to run on “ordinary” hardware.

Anyway, if your budget for the rest of your system is maxed out, by all means get a high-end pro card.


#9

I think you meant, ‘isn’t maxed out’.

I have read post from other users using higher end gaming oriented cards, and they are all disappointed, all because they got the nice cards because they thought it would give them a lot more performance and it didn’t. They were very unhappy, and it wouldn’t surprise me if they will move on to other software, because of bad experience. Overselling and underproviding is always a killer for businesses.

So it is very important to understand the limits the of the software and hardware and the relationship between the two, so its all optimized well. So that one is not dissappointed. the overall configuration should perform as expected.

Here is your context for this conversation, the key information is on page 13 of the article:
AMD Radeon R9 290 Review: Fast And $400, But Is It Consistent?

That 30 million number is just user review of the product, that can be found at newegg, and I have copy and pasted below:

Pros: *very quiet
*runs incredibly cool
*takes everything i throw at it even games on ultra (since this is a workstation card)
*cheaper than the w8000 and performs about the same.
*4 monitors!
Cons: price. is still rather expensive.

amd drivers still aren’t as great as nvidia

			Other Thoughts: So, first off. Im upgrading from an msi gtx 560 ti 2gb.

i use my pc for gaming and 3d software like zbrush and 3ds max.

i can now get easily 30 million polys in zbrush without even a hint of lag or lost frame rate. same with 3ds max.

gaming is just as fluid as what you get on a gtx 680 or a 7970

works
well with unreal development kit and cryengine as well. no visible
changes in performance from this card as opposed to a gaming card

pc specs
Processor: intel i7 2600k at 3.4 ghz w/ cm 212 evo
motherboard: asus p8z68 vpro-gen3
hd: intel 520 ssd 120gb x 2
ram: gskill ripjaws x 4x4gb at 1800
running 3 monitors - shane s.


Thus based the nebulous and unstable comments you have made, that I should tentatively expect similar results while using rhino 5 viewport.

I mean seriously people what does software run on anyway, sugar and coffee?


#10

Actually it’s from people with high-end Pro cards who you read here being disappointed they’re not getting more bang for their buck on a regular basis. There’s a flip side to the fact that Rhino doesn’t “need” one. But hey go right ahead and spend a ton of money on hardware based on random internet reviewers using completely unrelated software.


#11

So you are saying that you do not support your software? Or are you saying that Rhino will achieve similar results with cheaper hardware?


#12

What Jim is saying and what I was saying before is:

a) Rhino is designed to run on low spec hardware. “Pro” hardware may improve performance but most likely will not. Because
b) Rhino is not a poly-modeler and uses an entirely different display pipeline. Comparing it to 3ds Max, C4D, related benchmarks or any other polygon mesh based benchmark is useless.

As a side note: your quote refers to a user of Zbrush and Max. Both poly modelers. Both specifically designed to handle lots and lots of polygons and nothing else. Especially Zbrush generates a s*load of polygons while sculpting away. This user will no doubt see a better performance.


#13

mmmmm, McNeel is price gouging people with this product by lot! I would say its only worth about $150.00, not $900.00!

Ironically if you sold it at the the correct price, your sales might be able to skyrocket into more mainstream markets. Though unfortunately there is freeware available that does most everything and more that Rhino does, though it is geared toward games and marketing, Rhino in appearance anyway, is geared towards building and designing boats, and anything that is computationally similar to boats. Hobbyware is what it is.

Your total inability to clearly nail your product down to hardware, eliminates you from any professional interest, except I guess the narrow category of boat builders who seem to have been satisfied with this line of products for the last twenty years. No professional in his right mind would use this product, its nebulous hardware specifications are simply flabbergastingly absurd, there is no excuse for this, unless your market is simply not professionals.

the information you guys have given me certainly implies this, you are training students, but you are not keeping professionals, so you have a financial crunch, you can’t afford to support the product properly for the professional.

my interest in this product goes beyond common sense however. So I will keep using it for the time being. Though i should get at least a $700.00 credit and a free copy of bongo 2 for all the unGodly hassle you’ve guys have put me through. Seriously this is ridiculous. Their is no excuse for this runaround you are giving your customers who may have blown $1000’s of dollars to pump up some software that can’t be pumped up. Inexcusable. Highly misleading. Again and again I have given you guys specific information for professional interest, and if I had been hasty I would be even more unhappy and pissed off now than I am, having sold people on the a notion that can’t currently be fulfilled. performance information should be readily available for customers. simply ridiculous


#14

Heh, have we found an alt of JB here? I’m just waiting for the plug for SolidThinking…

I don’t know who’s giving you a runaround here, no actual McNeel staff have replied we’re just offering our opinion that you can save a few bucks. Anyway, I would hope “professionals” would investigate a little more what a video card does before dropping the value of a good used car on one, learn the difference between CAD and 3D graphics applications, and know better than to try to impress a board full of professionals by loudly claiming to be the mostest professionallist of all.


(Pascal Golay) #15

Hi Justin -
You can get a refund if you are not satisfied with Rhino.

-Pascal


#16

How long is that statement good for?


#17

I am a customer of the product. I have not bought one thing for this particular software. I’ve been digging around for information for over a month now, to select the correct parts, and its all very ambiguous.

To sum up:
----Powerful CPU with lots of GHz, probably don’t need more than 4 cores
----lots of speedy memory
----good motherboard
----SSD and HD
----and there is no good video card for this application currently as far as anyone knows, and no concise information about the subject anywhere that I can find, just ambiguity and outdated information from several years ago.


#18

Hi Justin,
I agree that Mcneel hasn’t provided comprehensive information about hardware performance information officially, which might lead Rhino users going wrong directions. I am one of the “victims” (I don’t really mean that) who bought expensive PC hardware that cannot take advantage of in Rhinoceros. But the conclusion of only professionals use Pro specs is a wrong statement. I’m an architecture STUDENT(but I do model industrial models for fun). I use Pro cards to improve the productivity. To me, serious people are Pros. I read a thread about similar article a while ago, and one Mcneel member stated that Rhinoceros is for all users who don’t have, or can’t afford decent computers. To be honest, I’m not rich, but I buy all licences, including Rhinoceros, Autodesks(almost expired), Adobe, Microsoft, Maxwell, Snagit and most likely will purchase and upgrade Rhinonest, Rhinoresrf and so on in the future. I can tell all my savings will be gone when I upgrade and purchase new licences. I wish,as many others do, Rhinoceros can take advantage of computer hardware.

Rhino is bad at mesh/polygon modeling but It’s good for Nurbs. Pro users don’t care what software they use.

Just let you know what Rhinoceros is capable of.

Class-A modeling in Rhinoceros. Download the file and Zebra the model.
https://www.dropbox.com/s/7j7zcm2taz98zlq/Class-A%20Modeling.3dm

-Kev


#19

Yes McNeel has really screwed this up, the marketing is so geared towards the ignorant newbs who don’t know much, which is me. People who can’t afford good computer will use free 3d software, or there price range is around $150-300 for it. there is great free 3d software available to people with little money. the only reason they would go with this stuff is because they did not know any better, newbs just gotta take a plunge somewhere. Or they are highly specialized, and it just does what they want it to do the best.

However, for reasons that make no common sense, i think the Rhino software package has a future and will prosper.

Whatever that file is, I can’t open it.


(Wim Dekeyser) #20

That link goes to page 12 so you’ll have to turn to the next page:

For CAD applications employing OpenGL, the Radeon R9 290’s performance is enough for a place in the middle of the pack, but that’s about it. The realized benchmark results are far from what’s theoretically possible based on the hardware’s specifications, but OpenGL performance remains largely dependent on driver optimization, which just isn’t a very high priority on gaming graphics cards. The situation changes a bit from one application to another, but OpenGL is certainly not the Radeon R9 290’s strong suit.

I can’t really see how you can be misreading that information? It’s right in your face: a gaming card, not for professional OpenGL use.

If you had been following the GPU thing, you’d seen that it is basically all down to the drivers. And they tend to be a mess. Instead of requiring expensive hardware and associated high fees for driver qualification at the card vendors, Rhino has attempted to use the “The Industry’s Foundation for High Performance Graphics” [quoted from www.opengl.org] as it is written (granted, an older version of it) so that it can be used on any hardware.

I don’t think anyone around here will try to stop you from spending a few thousands (every year) [unspecified currency;-)] on software that requires high-end cards. For the past 17 years I have been using Rhino side-by-side with SolidWorks, Pro/E, and now NX and I know which one I prefer…

By the way, an interesting (to me) comment on this review was on page 12 about DirectX:

DirectX’s disadvantage is its use of single-precision coordinates, which can easily lead to display errors in complex models […]