PDA

View Full Version : The LightWave 3D Group Welcomes Dominick Spina (formerly of nVidia) as V.P.



Chuck
08-03-2012, 02:23 PM
The announcement is here:

http://forums.newtek.com/showthread.php?t=129727

Discuss here! :)

OnlineRender
08-03-2012, 02:36 PM
congrats

Ztreem
08-03-2012, 02:50 PM
Congrats! Feels like it's just getting better and better,

MAUROCOR
08-03-2012, 03:00 PM
You have no idea how Lightwave 3D Group sounds good in my ears!:D

Yeah, I have to admit it is getting better and better. Welcome and Congratulations, Dominick Spina!

“The new LightWave 3D Group is elevating its global brand with new products, partnerships and development relationships that build on the tradition of innovation LightWave is known for,” explained Rob Powers, president, LightWave 3D Group."

What that means? "is elevating its global brand with new products..."

bobakabob
08-03-2012, 03:06 PM
You have no idea how Lightwave 3D Group sounds good in my ears!:D

Yeah, I have to admit it is getting better and better. Welcome and Congratulations, Dominick Spina!

“The new LightWave 3D Group is elevating its global brand with new products, partnerships and development relationships that build on the tradition of innovation LightWave is known for,” explained Rob Powers, president, LightWave 3D Group."

What that means? "is elevating its global brand with new products..."

Bring on yet more speculation ;)

shrox
08-03-2012, 03:18 PM
Are you bringing Stephen Hawking on too?

evolross
08-03-2012, 03:27 PM
Exciting news! Can't wait to hear what comes out at SIGGRAPH... I'm assuming it'll be the official announcement of LW11.5. Great stuff Newtek! :thumbsup:

nickdigital
08-03-2012, 03:28 PM
Great news! :thumbsup:

shrox
08-03-2012, 03:55 PM
Lightwave should be the app of the 22nd century it seems.

ivanze
08-03-2012, 04:10 PM
Lightwave should be the app of the 22nd century it seems.

No way!!! Iīll be dead by that time. :D

shrox
08-03-2012, 04:19 PM
No way!!! Iīll be dead by that time. :D
I meant the 22nd century, today!

cagey5
08-03-2012, 04:25 PM
Have we skipped a century when I wasn't looking?

shrox
08-03-2012, 04:47 PM
Have we skipped a century when I wasn't looking?

I am hoping Newtek has.

LW_Will
08-03-2012, 10:04 PM
Hmm... how much was that nVidia 670? Hmm...

Philbert
08-03-2012, 10:23 PM
Welcome Dominick!

kfinla
08-03-2012, 10:44 PM
Cool news, I'm looking forward to visiting the Newtek booth next week.

Verlon
08-03-2012, 11:14 PM
Like Lightwave hasn't had enough trouble with AMD/ATI cards already :devil:

Ok, seriously, great news. nVidia has never been short on talent. Hopefully this will lead to more exciting developments in Lightwave.

Silkrooster
08-03-2012, 11:28 PM
Congrats and welcome...

mav3rick
08-04-2012, 05:20 AM
welcome on board! now lets kick lw into GPU stratosphere :)

bazsa73
08-04-2012, 05:35 AM
GPU rendering? Perhaps my precious yes.

prometheus
08-04-2012, 06:26 AM
Lovely, seems like another human resource with strengths from the world leader of graphics card solutions.

Im quite happy with Nvidia graphics, and does not recommend anything else really when working with 3d.

Personally I enjoy the speed when dealing with simulating fluids (turbulenceFD) or using Fractron 9000 or using after effects.

Would like to see some GPU rendering options too in the future, both for a fluid interactive renderer, and for VPR perhaps.

perhaps Even bullet simulations?

However, I suppose Dominick will be in charge of mostly marketing thou, but It might help have his overall background from Nvidia anyway..

So Welcome Dominic, You will find that Lightwave has picked up a boost in development, and is definitly heading on to strengthen itīs software
much more nowadays, and perhaps even so the Market position.

Overall you should know that most people here in the community is very
exited nowaday with the direction of Lightwave.

Michael

COBRASoft
08-04-2012, 06:36 AM
Welcome! So, GPU experience to the table... Keep it coming!

KillMe
08-04-2012, 06:41 AM
all well and good but would like to see more people from the technical side of things being brought in too

50one
08-04-2012, 06:46 AM
This is a great news!:thumbsup:


all well and good but would like to see more people from the technical side of things being brought in too

I would like to see Vlado from Chaos Group, Eric Mootz and Thiago Costa within the Lightwave team hehehe:D

Amurrell
08-04-2012, 08:32 AM
Like Lightwave hasn't had enough trouble with AMD/ATI cards already :devil:...

Heh... I was thinking the same thing. :)

Congrats!

mav3rick
08-04-2012, 09:27 AM
never been fan of ATI (except for gaming consoles)

zapper1998
08-04-2012, 12:15 PM
Drats I wanted the Job...

He beat me to it..

Congrats... Dominick Spina..

:)

rcallicotte
08-04-2012, 01:08 PM
Congratulations Dominick! This is an exciting time to be a Lightwaver.

:lwicon:

antsj
08-04-2012, 04:15 PM
+1 to welcoming Dominick -
:thumbsup: to bringing GPU Rendering & CUDA to Light Wave.

aj

Philbert
08-04-2012, 04:57 PM
Not that I would know but since Dominick is VP of Sales and Marketing I'm not sure what he could do about Cuda or GPU rendering.

Oedo 808
08-04-2012, 07:04 PM
Not that I would know but since Dominick is VP of Sales and Marketing I'm not sure what he could do about Cuda or GPU rendering.

I'm hoping that it's not what you know, but who you know ;)

Welcome Dominick. :thumbsup:

Philbert
08-04-2012, 07:09 PM
In that case I hope you're right.

jwiede
08-04-2012, 07:34 PM
I'm hoping that it's not what you know, but who you know ;)
I think development resources are far more of a constraint there than lack of connections w.r.t. info. Actually implementing what you're asking for is a huge effort, but the knowledge of how to do so is quite available these days.

Welcome Dominick. :thumbsup:
Aye, welcome to Dominick Spina!

Oedo 808
08-04-2012, 11:24 PM
I'm not sure on Rob's appointment as president of the LightWave 3D group, I keep telling him that he should employ Google-savvy hobos and pay them with the odd portion of chips now and then, thus increasing available development resources. But he will insist on hiring people with experience and knowledge in the hope that it will somehow count toward something.




(BTW Congratulations on your promotion, Mr Powers).

shrox
08-05-2012, 01:56 PM
With this dream team we are seeing come together, might we get some specs for ideal Lightwave machines?

robertoortiz
08-05-2012, 02:40 PM
I do hope the people in NewTek will remmeber those who were with them
in the "dark times" that were the 00;s.


Great news guys.
-R

prometheus
08-05-2012, 05:29 PM
With this dream team we are seeing come together, might we get some specs for ideal Lightwave machines?

Highly dependent on your wallet.:D

Michael

prometheus
08-06-2012, 05:03 AM
Dominic would probably promote nvidia cards anytime wich is understandable, if we could get information from either Newtek dev teams or through Nvidia devs or former guys from there about different Nvidia series and how/if those take advantage of some Lightwave tasks, or plugin related, that would be great.

Probably You might be pointed to most expensive cards, It would be great to have an Honest recomendation without anyone just going for the selling
point of view, but with focus on what would be necessary and reasonable for performance.

I think Many of us understand that Nvidia cards are very good, but many of us do not know wich of the cards would be enough.

Extreme or large fluid simulations for example, will Geforce cards be enough for this, just making sure of huge memory and amount of cuda cores, or is it such a huge difference between latest quadro fermi cards for solving fluids.

Ivé been reading some topics regarding how quadro was so important for progress within cuda rendering and simulations.

Of course an article on the Nvidia site.

http://www.nvidia.com/object/the-last-airbender.html

COBRASoft
08-06-2012, 06:59 AM
It's quite easy for the moment. LW doesn't use the GPU, only CPU. LW uses OpenGL, not CUDA or OpenCL. So, the CPU is more important. I hope this will change in the future though. I suggest a Nvidia 2xx or higher. I use a 460 (desktop) and 675M (laptop) myself. But these or more for AE than LW.

prometheus
08-06-2012, 07:59 AM
It's quite easy for the moment. LW doesn't use the GPU, only CPU. LW uses OpenGL, not CUDA or OpenCL. So, the CPU is more important. I hope this will change in the future though. I suggest a Nvidia 2xx or higher. I use a 460 (desktop) and 675M (laptop) myself. But these or more for AE than LW.

Nope, not entirely correct, if you set up to work with turbulenceFD, wich of course works inside of Lightwave, You can take advantage of cuda, and its a Huge difference in simulation speed.

I just hope we get Jascha to write a dedicated turbulence previewer..with
Cuda rendering, I think Jascha mentioned in a discussion with me that He agreed that VPR isnīt ideal for turbulenceFD. cinema4d has its own previewer, not sure if it uses cuda thou.

I think we need a previewer able to handle multiscattering, wich you can not see in VPR while you are tweaking the fire and smoke shader, but you will see in final render, so I am all for an optimized cuda renderer or previewer for turbulence, jascha said he Might look in to this, but couldnīt promize anything.

I do have suspicious thoughts thou about the Bullet engine, I would not be surprised if we will see the engine taking advantage of cuda very soon.

Michael
Michael

COBRASoft
08-06-2012, 10:33 AM
Prometheus: sorry, I don't agree. Lightwave doesn't use CUDA or the abilities of advanced GFX cards. If you have no need for Turbulence3D, you don't have to buy a heavy CUDA enabled GFX card. Also, a NVidia 2xx or up should do fine with Turbulence3D. I'm sure more plugins will come that use the GPU. But TODAY, it's very limited inside LW. I'm still convinced that buying a 3rd generation I7 will do more miracles to LW than any GFX card.

prometheus
08-06-2012, 10:54 AM
Prometheus: sorry, I don't agree. Lightwave doesn't use CUDA or the abilities of advanced GFX cards. If you have no need for Turbulence3D, you don't have to buy a heavy CUDA enabled GFX card. Also, a NVidia 2xx or up should do fine with Turbulence3D. I'm sure more plugins will come that use the GPU. But TODAY, it's very limited inside LW. I'm still convinced that buying a 3rd generation I7 will do more miracles to LW than any GFX card.

Uhmm..I donīt follow you,?why narrow down to Lightwave doesnt use cuda, Turbulence works within the framework of cuda does it not?
and fairly it can be said taking advantage of cuda, I do not see a reason
to separate two tools that are dependent on each other to work.
Thatīs why I wouldnīt say cpu is important and cuda not, depends on what you plan to work with as a starter, do not rule out fluids as a tool generally
for people, they need to know the facts around that.

And yes it is very limited, but the point was you said that it doesnīt use cuda when in fact it does with TurbulenceFD, so how can we disagree on that?:stumped:

Besides, why focus on Lightwave only? I have use for it with fractron render9000 (free software), octan render, adobe after effects, and I think Itīs just a matter short time before it shows up in Lightwave native, probably within the Bullet engine.

when thinking on buying graphics cards you should also take in consider some other toolset enhanced by this, like 3d coat perhaps, but if youre only tool is lightwave and no special future plans, sure go with something else.

Michael

prometheus
08-06-2012, 11:03 AM
Besides, I think you are a little wrong on about advanced graphics cards, that is a little vague isnīt it, think about handling large polyobjects wich in fact I am certain of that quadro cards will deal with better than many geforce series. they are mostly equiped with more memory and enhanced drives for performance of that.

graphic cards are important, check Dj lithiums youtube videos on handling of huge spaceships etc.

Michael

COBRASoft
08-06-2012, 11:22 AM
Prometheus: if you read my first comment again, then you'll see that I have quite an advanced GFX cards for other apps (AE). So yes, for those it does really matter. For high polycount you need more memory, not the latest GFX card out there. I've done a lot with my 260 last year with 2GB ram on it. I also wrote that I'm talking about Lightwave TODAY. I sure hope it will change and that the GPU becomes more used in whole LW. We can only hope that Rob and his team give us a glimps of the future concerning GPU and LW.

Quadro cards... LW doesn't use any other special ability of those cards if it doesn't come standard with the driver. The latest quadro drivers from NT go back ages (http://www.nvidia.co.uk/page/ws_partner_certified_drivers.html). There have been threads about Quadro cards before and the result was each time the same. The price doesn't pay off the small benefits within LW world.

To everyone: if I'm talking bullsh*t here, please correct me. The last thing I want is to misinform people.

Dexter2999
08-06-2012, 12:08 PM
I think what it comes down to is what will the manufacturer recommend for their product. And their product doesn't use the graphics card (as of now) for processing.

Expecting Newtek to make recommendations based on plug ins or applications you may or may not have is ill advised.

If Turbulence is what is using the graphics card, then look to its makers for recommendations about a graphics card, not Newtek. Similarly with other programs.

shrox
08-06-2012, 12:12 PM
It's quite easy for the moment. LW doesn't use the GPU, only CPU. LW uses OpenGL, not CUDA or OpenCL. So, the CPU is more important. I hope this will change in the future though. I suggest a Nvidia 2xx or higher. I use a 460 (desktop) and 675M (laptop) myself. But these or more for AE than LW.

Right there is something I did not think about, it would be nice to have some more useful info as well. That is why I would like see some specs, maybe even based on budget, what one saves in a video card (since the GPU is not used) one could put towards ram or a faster CPU.

prometheus
08-06-2012, 01:07 PM
Prometheus: if you read my first comment again, then you'll see that I have quite an advanced GFX cards for other apps (AE). So yes, for those it does really matter. For high polycount you need more memory, not the latest GFX card out there. I've done a lot with my 260 last year with 2GB ram on it. I also wrote that I'm talking about Lightwave TODAY. I sure hope it will change and that the GPU becomes more used in whole LW. We can only hope that Rob and his team give us a glimps of the future concerning GPU and LW.

Quadro cards... LW doesn't use any other special ability of those cards if it doesn't come standard with the driver. The latest quadro drivers from NT go back ages (http://www.nvidia.co.uk/page/ws_partner_certified_drivers.html). There have been threads about Quadro cards before and the result was each time the same. The price doesn't pay off the small benefits within LW world.

To everyone: if I'm talking bullsh*t here, please correct me. The last thing I want is to misinform people.

Youre telling me that latest quadro drivers go back ages? and that is how long? as to what I see on their drivers site there are updates 2012-07-18 and that is ages ago? what are you talking about?

The best one with insight on this could perhaps be Dominik Spina himself telling if quadros outperforms geforce cards in handling extreme polyamount and open gl performance.

Neither you or I seem to have experience with higher end latest quadro fermi cards, so I think none of us really can give the proper truth on whatīs correct here.

I had a quadro 1100 before and was quite impressed with handling large data with that, and the machine on board ram was much lesser than what I have today in my desktop specs as you can see in my signature geforce 480 card.

Havent had any chance to try and compare since that old computer is down for quite some time and I need to replace stuff in there If Im up to it, otherwise I am using my laptop for the moment..

8gb ram geforce gtx 560M
Asus intel i7-2670QM cpu 2,20 GHz

Can we agree on that you donīt think lightwave can be said to use cuda, since itīs only the plugin using it and thatīs how you see it? while I think it does since it is part of the whole working framework, but we differ here?

That was my only objection about your statement wether lightwave uses cuda or not.

Then we could go on discussing what will happen and what is, I just cant lock myself to a static today, I do not go out and buy a computer or software solely on what it can acheive today, and today is tomorrows yesterday, and soon lw 11.5 is upon us, who knows if they have or will implement further support for cuda when that arrives.

Michael

prometheus
08-06-2012, 01:12 PM
For the getting the record straight, buy anything less than cuda supported cards, and you can not use GPU enhanced improved speed for the simulations. this isnīt disputed.

with turbulence you have an option to choose either cpu, or gpu if you have cuda supported cards that is.

Michael

shrox
08-06-2012, 01:14 PM
Some laymen info about the actual hardware usage of Lightwave would be cool too.

prometheus
08-06-2012, 01:19 PM
I think what it comes down to is what will the manufacturer recommend for their product. And their product doesn't use the graphics card (as of now) for processing.

Expecting Newtek to make recommendations based on plug ins or applications you may or may not have is ill advised.

If Turbulence is what is using the graphics card, then look to its makers for recommendations about a graphics card, not Newtek. Similarly with other programs.

True I guess, they can however have an extensive experience of third party and some knowledge of whatīs out there, otherwise what the heck are they doing?

so giving a general advice on a good card working for Lightwave can also extend to be a good card handling many required task for third party tools to work properly within Lightwave.

But Agreed, generally it should be adviced to research the potential third party developers site.

prometheus
08-06-2012, 01:34 PM
Volumedic..

I do not think Elmar has worked on support with cuda for his GPU raycast
and open GL shading thou, I think I read somewhere that He would never add support for that, canīt say for sure if that was the case, and what dislikes he had about it?


Michael

DogBoy
08-06-2012, 03:44 PM
Youre telling me that latest quadro drivers go back ages? and that is how long? as to what I see on their drivers site there are updates 2012-07-18 and that is ages ago? what are you talking about?

I think he means the last time Quadro drivers had an update that addressed LightWave changes was a long time ago. LightWave doesn't, as a rule, push OGL in ways that some of the other apps do. If memory serves only recently has NT defined the minimum specced vid card for decent performance.

prometheus
08-06-2012, 04:49 PM
I think he means the last time Quadro drivers had an update that addressed LightWave changes was a long time ago. LightWave doesn't, as a rule, push OGL in ways that some of the other apps do. If memory serves only recently has NT defined the minimum specced vid card for decent performance.

Ahh..I see the point, but it confuses me that updates of gtx drivers has been necessary in order to have TurbulenceFD working properly, so the argument doesnīt hold up anyway ...I think.

Michael

COBRASoft
08-06-2012, 05:57 PM
Prometheus: please read my answers a little more detailed :). LW didn't provide (agreed) a certified driver for Quadro cards for years. NVidia has much newer drivers, but the user has to search for the best settings for Lightwave. Even worse, Lightwave is not even in the preset list for settings (Spina, please use your connections to correct this). My latest quadro was a 3xxx in a HP workstation. So, quite some experience (and problems with LW).

About LW and CUDA. LW's code doesn't have CUDA integrated. The coder of Turbulence3D has to implement this on his own. Turbulence3D uses parts of LW code (exchange data and so on).

As for buying a new computer... As the main software developer of an industry leading ERP app (in Belgium), digital artist and other stuff, I have VERY high demands of my computers. I try to 'forsee' 1 year in the computer world before buying one.
Will LW 11.5 have CUDA ready to use for the users? I don't think so. LW12 maybe, but if it has, it will start as a basic implementation IMHO. Don't forget that LW supports ATI also. ATI doesn't have CUDA.

Side note: I hope I'm wrong with my assumptions :).

prometheus
08-06-2012, 06:20 PM
Prometheus: please read my answers a little more detailed :). LW didn't provide (agreed) a certified driver for Quadro cards for years. NVidia has much newer drivers, but the user has to search for the best settings for Lightwave. Even worse, Lightwave is not even in the preset list for settings (Spina, please use your connections to correct this). My latest quadro was a 3xxx in a HP workstation. So, quite some experience (and problems with LW).

About LW and CUDA. LW's code doesn't have CUDA integrated. The coder of Turbulence3D has to implement this on his own. Turbulence3D uses parts of LW code (exchange data and so on).

As for buying a new computer... As the main software developer of an industry leading ERP app (in Belgium), digital artist and other stuff, I have VERY high demands of my computers. I try to 'forsee' 1 year in the computer world before buying one.
Will LW 11.5 have CUDA ready to use for the users? I don't think so. LW12 maybe, but if it has, it will start as a basic implementation IMHO. Don't forget that LW supports ATI also. ATI doesn't have CUDA.

Side note: I hope I'm wrong with my assumptions :).


I will take a look in to reading your answers with a little more focus on the details.


Lightwave has no cuda intergrated, that hasnīt been disputed and I havenīt said so, probably missunderstanding in communication when I mention Lightwave and Turbulence, Havenīt really had the need to in discussions dissect it t seperate the two of them, since they are dependend on each other to do the job, the fact that Jascha has
developed cuda of his own is very correct, I also think that it might not show up until lw 12, if it does?

I however found it a little pointless of you to debate the true nature of wether lightwave uses cuda or if it should be called Lightwave together with Turbulence take advantages of cuda, or turbulence takes advantage of cuda and so on, sure you can pin point it down to exact nature, but getting there to sort it out with such a discussion feels so unnecessary if you exuse me of how I feel about it, if anyone mentions Lightwave and not taking advantage of cuda, that could easily fool anyone to disregard what can be acheived with it together with plugins That for real take advantage of it.

Just thinking, these explanations you post with all knowledge described in later post, why couldnīt you have mentioned simply straight to the point in first posts, like
lightwave doesnīt use cuda in itīs code, turbulenceFD is special coded for it, etc.

took some time to get there, my conclusions is just that you canīt get the benifit of performance of fluids in lightwave without Cuda supported cards.

Dog tired now, have to sleep a few hours and up to work, beginning to loose my english spelling too now.

prometheus
08-06-2012, 06:28 PM
ABout the preset and quadro drivers, isnīt quadro the only nvidia drivers, Newtek say is fully supported to use with lightwave?
and geforce series is not?

DogBoy
08-07-2012, 03:22 AM
ABout the preset and quadro drivers, isnīt quadro the only nvidia drivers, Newtek say is fully supported to use with lightwave?
and geforce series is not?

Newtek have found no reason to update the Quadro driver needs. You can use the newest and it will be fine. Geforces offer exactly the same functionality for LightWave as a Quadro, which is why everyone says save your money if you don't use more Quadro needy apps, like Maya, Max, Soft, Mudbox ......

As for presets, well to be honest I've never cared. Sure, you could tweak your profile to get the most out of LW, but frankly there isn't much point as out of the box it works pretty well.

prometheus
08-07-2012, 05:19 AM
Newtek have found no reason to update the Quadro driver needs. You can use the newest and it will be fine. Geforces offer exactly the same functionality for LightWave as a Quadro, which is why everyone says save your money if you don't use more Quadro needy apps, like Maya, Max, Soft, Mudbox ......

As for presets, well to be honest I've never cared. Sure, you could tweak your profile to get the most out of LW, but frankly there isn't much point as out of the box it works pretty well.

Yes, Ivé heard that a lot, but I have a hard time grasping that as the truth, even though so many speaks up about that, donīt know why Iīm having a hard time accepting that. maybe Im getting fooled by technical specifications of the higher quadro cards that in fact has more on board graphics memory and so on, and therefore I can not understand why that
wouldnt work more efficient in handling huge objects within lightwave?

Can anyone give a short explanation about the architecture within Maya
and the difference against Lightwave when it comes to what the quadros handles well as a complementary explanation rather than just ..it is needy app?



Thanks

Michael

DogBoy
08-07-2012, 07:12 AM
My (uneducated) take is that there are 2 areas of difference: that due to drivers nVidia have made to differentiate between Geforce and Quadro, and Newteks OpenGL implementation.

Newtek:
Newtek has traditionally gone for OpenGL for 3D acceleration, but due to Apples very slow adoption of newer releases of OGL, NT never really utilised a lot of OGLs newer features.

OpenGL can do most of the stuff DX11 can do, including hardware tesselation. If NT coded for that, we wouldn't notice the slow downs when we upped Dispay Subpatch Level, as it would offload that to the video card and not do it "in house". This also why people complain about LWs poly-handling, while other apps seem much better: they offload a lot more of the display handling to hardware, where NT do it in software.

This is why we had the whole "drop OGL and move to DX" arguments of a few years back. LW users believed Apples crap OGL support was holding back LW. They felt if NT adopted DirectX LW would improve. They were partially right, but those improvements could be got from OpenGL if Apple had just pulled their finger out.

Newtek added GLSL textures in LW9 (i think), but had to use an earlier implementation, so the benefits were marginal compared to what you would see in some thing like Mudbox or Mari.


Nvidia
nVidia felt the need to turn off a bunch of features that they felt consumer level cards didn't need.
Clip Regions: They allowed windowing/clip regions of OpenGL acceleration on Quadros only. This means if you have a a non-3D window overlaps a 3D window (say Graph Editor sitting over your Viewport) The Quadro doesn't have to compute the parts of the viewport that can't be seen, so they get better performance.
Windowed Multi-Card Acceleration: GeForce cards can finally run SLI mode on multiple monitors, which has been available on Quadros for a few years. What they can't do is do it in a Window, so it doesn't work with applications like LW where you only want to do it in the Viewport.

Add to that things like Line AA (LWs grid AA is software) & Overlay Planes (to draw those funky tool gizmos, Maya and most other 3D apps use hardware accel), and the Quadros benefit apps coded to use them.
Plus let's not forget better colour output; Quadros do 10bit unlike Geforces.

COBRASoft
08-07-2012, 07:16 AM
I know for a fact that quadro cards are a lot better in handling overlapping windows. May seem a stupid detail though...

DogBoy
08-07-2012, 07:21 AM
I know for a fact that quadro cards are a lot better in handling overlapping windows. May seem a stupid detail though...

That's what i meant about Clip-Regions, you just said it clearer :thumbsup:.

I found this link (http://www.google.com/url?sa=t&source=web&cd=1&sqi=2&ved=0CBMQFjAA&url=http%3A%2F%2Fwww.nvidia.com%2Fobject%2Fquadro_ geforce.html&rct=j&q=geforce%20vs%20quadro&ei=p5DhTMPdGISs8AaXr_ED&usg=AFQjCNESkCnpBGLD_XkAyUuFJ4Ky6W-D_g&cad=rja) to an old nVidia paper. It is old, but mostly cover the principles. It doesn't cover SLI so no mention of multi-card multi-monitor support or 10bit colour.

Wikipedias take:


The Quadro line of GPU cards emerged in an effort at market segmentation by NVIDIA. In introducing Quadro, NVIDIA was able to charge a premium for essentially the same graphics hardware in professional markets, and direct resources to properly serve the needs of those markets. To differentiate their offerings, NVIDIA used driver software and firmware to selectively enable features vital to segments of the workstation market; e.g., high performance anti-aliased lines and two-sided lighting were reserved for the Quadro product. In addition, improved support through a certified driver program was put in place. These features were of little value in the gaming markets that NVIDIA's products already sold to, but prevented high end customers from using the less expensive products. This practice continues even today although some products use higher capacity and faster memory.

creacon
08-07-2012, 07:36 AM
I'm using PhysX for my plugin and can't seem to get a GeForce to simulate in double precision mode. I wonder if that is one of the limitations.

creacon


I know for a fact that quadro cards are a lot better in handling overlapping windows. May seem a stupid detail though...

prometheus
08-07-2012, 07:52 AM
I found an interesting Swedish article about the comparison of quadro cards and gtx cards, you would have to translate through google and some might be lost in translation thou.

some are mentioning that the memory ECC that accounts for high accuracy when calculating decimals and Thatīs more stable and important for crucial simulations or something? and that ECC is only available in quadro cards,

some mentioned that in lower end quadro cards, and gtx? there as been applied a sort of speed limit? for calculations of decimal numbers with high accuracy, that is double precison float (64 bit)
and also the fact that for instance quadro fx 6000 has 6 gb Vram compared to 470 which the quadro card is based on?

Which means that very expensive higher end cards should outperform most gtx cards for ..letīs say fluid simulation?

Also mentioning 10 bit color depth per Rgb channel which is supposed to be important for some visualization areas.

I just tried to translate a little here but..if interested check yourself and see what you think, I recognize the other stuff you guys mentioned about differences from some time ago, I had an old quadro fx 1100 card, the computer broke, and I got me one with gtx 480 instead.

makes me wonder thou that ILM proffessionals set up high end quadro cards to handle massive fluid effects, even thou they use their own fluid
proprieatary tool called plume, so why can they not settle with less? they arenīt doing solidworks exactly.

Hereīs the article pointed to google translate also.
http://translate.google.com/translate?sl=sv&tl=en&js=n&prev=_t&hl=sv&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.sweclockers.com%2Fforum%2F125-geforce%2F982342-geforce-vs-quadro-varfor-denna-sinnessjuka-prisskillnad-och-behovs-quadro%2F&act=url

Cheers, this is informative anyway.

COBRASoft
08-07-2012, 11:20 AM
ECC memory is very expensive compared to normal memory. If you want a GFX card with ECC memory, then you should place ECC memory in your computer too IMHO. My HP workstation had this ECC memory and costed me a fortune :).

I guess it all depends for what you will use your computer. If fluid anims or turbulence3D is used daily, go for high-end. If you're more of an archviz person or so, go for mid-range. Quadro is only needed for specific programs (e.g. AE6 raytracer). If you use those a lot, go deep in your pocket...

Nevertheless, I stay by my first argument. Pay for a better CPU like the 3rd generation I7. ALL your programs will benefit from it, specially when you're running them at the same time. Oh yes, also go for SSD drives.

prometheus
08-07-2012, 11:18 PM
ECC memory is very expensive compared to normal memory. If you want a GFX card with ECC memory, then you should place ECC memory in your computer too IMHO. My HP workstation had this ECC memory and costed me a fortune :).

I guess it all depends for what you will use your computer. If fluid anims or turbulence3D is used daily, go for high-end. If you're more of an archviz person or so, go for mid-range. Quadro is only needed for specific programs (e.g. AE6 raytracer). If you use those a lot, go deep in your pocket...

Nevertheless, I stay by my first argument. Pay for a better CPU like the 3rd generation I7. ALL your programs will benefit from it, specially when you're running them at the same time. Oh yes, also go for SSD drives.


Well fluids will probably be a big part of stuff to do, so a card that can handle that will be important, but it might not justify a higher end card anyway, at least not for a non studio guy, but depends on how big the difference really is also and if you really are in need for it during work projects.

solid state..sure, but the difference between the cost and value is still to much today. 3 times more expensive than 7400 cards, and you only get half the storage if even that.

Michael