PDA

View Full Version : Difference of game 3D card and graphic 3d card.


dickma
11-14-2003, 03:55 PM
Can someone state the difference of game 3D card and graphic 3d card in performance of running 3D software?

From Tom's hardware guide I see the Quadro 3000 and the Geforce 5900 has the same GPU. So, what is the difference in performance?

I am confused. or someone have an experience before and after changing card?

Thalaxis
11-14-2003, 04:40 PM
The difference is mostly in the drivers. There are some features
that they do not enable in the GeForce version that they do
enable in the Quadro version, like anti-aliased line-drawing. In
the GeForce version, that is handled in software, while in the
Quadro version that is handled in hardware.

It's for market segmentation, just like AMD's not allowing AthlonFX
to run in SMP environments, so that they can charge a higher
premium to markets where the premiums are not as important as
that extra capability, so they don't mind paying it.

dickma
11-14-2003, 04:55 PM
How's important about the anti-aliasing line drawing, or other enhance functions, does it affect or enhance the production pipeline? Like in animation previewing, does Quadro faster than Geforce.

I heard somebody try to unlock the Geforce to Quadro. If I can do that it will be good because I can save a lot of $.

Actually I don't like those marketing segmentation, that cannot benefit the CG industry. Because some cards are too expensive that the designers cannot affordable.

Thalaxis
11-14-2003, 05:12 PM
Originally posted by Dick Ma
How's important about the anti-aliasing line drawing, or other enhance functions, does it affect or enhance the production pipeline? Like in animation previewing, does Quadro faster than Geforce.


The biggest difference is in drawing wire-frames. There are some
others, that's just the most noticable one from the user's
perspective.


I heard somebody try to unlock the Geforce to Quadro. If I can do that it will be good because I can save a lot of $.


They were in previous generations identical silicon, so it wasn't
much of a hack. It's still doable, but it ends up splitting the
difference now, rather than turning your GeForce into a full-
fledged Quadro.


Actually I don't like those marketing segmentation, that cannot benefit the CG industry. Because some cards are too expensive that the designers cannot affordable.

It's good for the vendors. They choose prices the market will bear,
and if the market will bear a higher price, it means more revenue
for the vendor. That's the ONLY reason that Xeon processors cost
more than Pentium4 processors, Quadros cost more than
GeForce's, FireGLs cost more than Radeons, and so one.

If you can't afford a Quadro, don't sweat it... just get a GeForce or
a Radeon, crank it up, and be happy when you find out how fast
it actually is. With most modern graphics hardware, you'll be
pleasantly surprised.

dickma
11-15-2003, 02:18 AM
It's good for the vendors. They choose prices the market will bear,
and if the market will bear a higher price, it means more revenue
for the vendor. That's the ONLY reason that Xeon processors cost
more than Pentium4 processors, Quadros cost more than
GeForce's, FireGLs cost more than Radeons, and so one.

If you can't afford a Quadro, don't sweat it... just get a GeForce or
a Radeon, crank it up, and be happy when you find out how fast
it actually is. With most modern graphics hardware, you'll be
pleasantly surprised.

It's too much like economics (I hate calculate!) and as you say. it may be just benefit the vendors, not the designers.

I may be affordable to buy a Quadro, but just feel miserable if I choose between a Quadro FX 1000 or Geforce FX 5900+. The Quadro FX is still double price of the Geforce FX one, but the Geforce FX use the same GPU that the Quadro FX 3000 using. :shrug:

CgFX
11-15-2003, 05:52 AM
First, they are different chips. FX 5900 is NV35 and FX 3000 is NV35GL. They share the same core architecture but are different chips (unlike GeForce2/Quadro2). As noted, there is driver magic as well to go with the extra silicon.

Second, we are paying $1500 for FX 3000 which is nearly 10 times the geometry performance of a Onyx2 InfiniteReality3 (avg. of about $200K to $500K). It is 3 times the pixel fill performance of the 4 RM ~$750,000 version.

Instead of bitching that the workstation version is so much more expensive than the gaming version we should be praising all the gamers for making the workstation board so cheap with all the volume of the core architecture they buy.

Or, we could go back to the way it was where there was no gamer version so you paid $3,000 for the Wildcat board that was from a company that ends up going out of business since only a handful of people would pay for that when it wasn't any faster than a Quadro4 900 XGL or wasn't worth 6x a Quadro2. And we can all see how involved in graphics SGI is anymore.

CGTalk Moderation
01-16-2006, 05:00 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.