Difference between softmodded & real QUADROs. And how do i know i match the model ?

:dave: what's up JEDIs?

I want to know, what's the difference between a softmoded quadro and a real quadro.

And how to corroborate that my softmoded quadro, match the capabilities of that model.

I feel like benchmarks apps doesn't really tell me anything.

I want to know if there's anything more tangible, like a render test or something that i can't do with traditional geforce, but that i could be able to do after the softmoding.


Sorry about my english.

Thanks in advance.

 

It is different for every program and how a quadro could benifit.

What program do you use?

 

I'm doing a small training in modeling. ( Maya ).

So far, i'm doing ok, cause we don't need to much in class. We're just modeling and working in wireframe mode.

But i really really wanna do some rendering. So i want to know about something tangible that i will be able to do in quadro mode.

Something, that i can't do with my geforce right now, at stock.

I don't know if i explain my self.

 

In maya:
Geforce
Paint effects will leave artifacts on the screen (basicly i believe it is a refresh sync problem, but other say it is overlay issue, but i have the same issue in another program that doesn't use overlays that i know of).

Quadro:
No artifacts

Also note, i here from alot of Maya users on Cgtalk that maya really is not that much faster with a quadro, but it does fix the visual issues.

Also just a note for you, the normal rendering system uses CPU power only not GPU, unless you are doing hardware rendering (maya has this option i think) it isn't a good quality render so it isn't really used much for final frames, software renderings (99% of all of them) are the best looking ones and they are defaulted and all programs use them.

 

cpu does rendering

hi
as said, in 3d apps like maya/max the quadro graphics card is for
moving your wireframe models (with a quad 4000 even shaded)
and you cpu/ram deals with the rendering.
i use 3dmax daily and this forums been a real eye opener.
ive got a real fx2000 (cheap from e bay) but after been here brought a
6800gt and my softmodded to quadrofx4000
beats fx2000 no prob.
i would never pay top doller for the real name on the card.
£1400 vs £160 when they seem to be the same thing
the modded 6800gt to quad is great and if your using 3d apps its
a must have cheap way to go.
all the best
dmb

 

Yeah but now you are limited at best a quadrofx4000... which if you game + do 3D apps, is not good enough with a 19x12 res :bolt: .

I haven't been doing anything to hardware in 3D since i got my 7800 but it seems to work pretty good in most of my apps but Rhino3D where i have to use software instead of opengl... but it is still very fast lol.

Maybe now that it seems impossible to find a way to use NVStrap to make the new Nvidia cards into Quadros maybe we will have to go back to modding the drivers.. if it hasn't be already thought about... (i am sure it has and no luck) :frown:

 

if you mod a 6800 Ultra it will be like having a Quadro FX 4400, since the clocks on the real Quadro FX 4000 are much lower (325/1000 vs 425/1100) and it uses 12p,6vp vs 16p,6vp on the 6800 Ultra.

And here are the differences between GeForce and Quadro i have been observing in pro 3D apps:

Cinema4D = no difference whatsoever, GeForce 6600GT works same as Quadro FX 4500
Maya = faster antialiased wireframes (soft wireframes), OpenGL hardware overlays in Artisan (= no artifacts in PaintFX)
3DSMax = if using Maxtreme drivers, the performance improvements in wireframe displays are huge
Houdini = faster performance all arround, due to heavy use of OpenGL hardware overlays
Gelato = GeForce FX 5900 works the same as Quadro FX 4400, since although Gelato is using GPU hardware for computation a lot, it seems that it depends on the CPU much more...

 

Hi, man. This stuff has already been discussed, you know. Quadros are much more expensive because it costs a lot to nVidia to create drivers. So instead of making people pay the drivers, they found a trick to make us pay the same things more expensive.

 

Come on guys, don't make this thread a battlefield. :dave:

This is sort of the answers i'm hoping to get:

:invasion:

 

Whaaat..??
I can understand the lowered clock speeds for ensuring stability but why did they make the retail Quadro FX 4000 to have 4 of the 16 pixel units blocked?
I mean, it was the top-of-the-line Quadro when it came and everything..

 

I suspect low yields on the first NV40 chips, requrement for silent operation (remember, it's meant for workstations ) and total irellevance of the number of pixel pipelines to the professional apps since they mainly process geometry - NVIDIA probably thought that having the maximum number of vertex processors (6) is enough. Also, if the Quadro FX 4000 had 16p,6vp and high clocks, it will be equal to performance to Quadro FX 4400 since the only difference would be the bus type (AGP vs PCIe) and we know that cannot be a reason for a huge modelnumber (and price) change...

 

Aah.. hehe.. yes, I understand now..!

 

What are advantages of professional apps in ProEngineer?

I have noticed in PTC ProEngineer, that although specview 8.1 shows a remarkable difference between a geforce and a quadro, in actual use I cannot tell the difference between a quadro fx3450, a 6800 ultra, or a 6800 ultra softmodded to quadro, but all are significantly faster than a quadro fx1300. I use fairly large assemblies, and have tried comparisons in shaded and wireframe (quite difficult since it takes about 20s for the regeneration of the model say after stopping spinning). Does anyone know what actual benefits ProEngineer sees from 'enabling professional apps' or using a quadro over say just a really fast card like a geforce 7800gtx or ati x1900xtx?

thanks

 

In the old (September 16, 2003) "NVIDIA Quadro vs. GeForce: Features and Benefits" technical brief
http://www.nvidia.com/object/quadro_geforce.html
Pro/ENGINEEER is mentioned at least in both the "Antialiased Points and Lines", the "Clip Regions" and the "Hardware-Accelerated Clip Planes" sections.

 

Thanks clokkevi
Excellent brief. I may even learn some more about the functionality of ProEngineer...I've only been using it for 10 years lol.