Jump to content
3DCoat Forums
L'Ancien Regime

GTX 780ti 6gb DDR5 RAM

Recommended Posts

26a.jpg

 

 

EVGA Rep also confirmed that there are no GTX 780 (Ti) with original blower type coolers planned. The standard GTX 780 is probably going to cost around $549.99 (roughly $50 more than 3GB models).

 

 

Share this post


Link to post
Share on other sites

$550 is rally cheap for this monster. However I'm more interested in how much will it cost in Europe. Probably at least twice as much.

Is this the planned retail price?

Share this post


Link to post
Share on other sites

wow... 3 years and we finally get GTX 580 CUDA performance back...  except at double the price.

 

still nice that we are back though... =))

 

I have waay too much hardware to bring back with me from my trip to new york :)

Share this post


Link to post
Share on other sites

wow... 3 years and we finally get GTX 580 CUDA performance back...  except at double the price.

 

still nice that we are back though... =))

 

I have waay too much hardware to bring back with me from my trip to new york :)

I agree on the CUDA side of things, but this has a 6GB framebuffer (VRAM), making it easier to render larger, more complex scenes in a GPU renderer, like Cycles, Thea, Octane, Arion, Moskito, Furryball, Redshift, etc. Plus, the card itself should be a decent step ahead of the 580 in overall mesh handling and scene navigation. The GTX 600-770 series were bottle-necked with a reduced Memory bandwidth (256bit). But with the 780, it went back up to 384. So, it's a good investment for anyone using a GPU renderer.

Share this post


Link to post
Share on other sites

$550 is rally cheap for this monster. However I'm more interested in how much will it cost in Europe. Probably at least twice as much.

Is this the planned retail price?

But its not the "Ti " version , as far as i can see. ??? The 780 ti usually costs around 650 dollars

Share this post


Link to post
Share on other sites

I agree on the CUDA side of things, but this has a 6GB framebuffer (VRAM), making it easier to render larger, more complex scenes in a GPU renderer, like Cycles, Thea, Octane, Arion, Moskito, Furryball, Redshift, etc. Plus, the card itself should be a decent step ahead of the 580 in overall mesh handling and scene navigation. The GTX 600-770 series were bottle-necked with a reduced Memory bandwidth (256bit). But with the 780, it went back up to 384. So, it's a good investment for anyone using a GPU renderer.

I think it's because Nvidia got hit pretty hard with the 400-500 series being such power hogs and running too hot.  I think they over-reacted with the 600 series and performance got hit really hard.  This is why I'm still using my two GTX460's, they are almost equal to a GTX680 in raw speed.  To give Nvidia some credit the 600 series(Kepler) has more texture space, they definitely run cooler, and use less power.  I'm keeping an eye on the 780's to upgrade to, because if I'm going to get any kind of speed bump, that is the level I need to go to.  $550 is a nice price.  :D

Share this post


Link to post
Share on other sites

It may be a bit over the top, but I went ahead and posted the video comparison I did for Andrew, on Youtube...to kind of warn others to do lots of research before they throw down hundreds of $$$ on a new Nvidia card. I got suckered, and ended up having to sell the card on eBay, for about a $50-75 loss....and just bought a used GTX 580 3GB instead.

 

https://www.youtube.com/watch?v=TU-LXFhCS_Y

Share this post


Link to post
Share on other sites

Damn stuff like this should be stickied to the top of ever software developers website..

Share this post


Link to post
Share on other sites

Damn stuff like this should be stickied to the top of ever software developers website..

Yeah. I've been checking out some of the leaked specs for the upcoming GTX 880. And I'll be a monkey's uncle if they didn't go right back down to a 256bit memory bus!!! What in the hell is going through their minds? I suspect it's a cost-cutting move. The memory bus lanes are probably using gold (for maximum conductivity), and that is why they are trying to skimp. Skimp on the budget models, but not on the higher end ones. If it costs an extra $50 to stay with 384bit bus, or go up to 512, then so be it. I'm not buying another graphics card with a 256bit bus. There is just no sense in it.

 

So, with that in mind, I guess one would best get a 780 or Titan and use that for the next 3-4yrs until NVidia gets a clue.

Share this post


Link to post
Share on other sites

I think Nvidia takes features out of the gaming cards in places where they believe the gamers won't notice.

Unfortunately there are a bunch of freelance 3D artists who use these gaming cards for digital content creation, and I have a feeling those artists are in the minority.

The big 3D studios still use render farms which are mostly CPU based. If the big studios still think they might possibly need more GPU power, then they will buy the top end Quadro cards.

That leaves the freelance 3D artists stuck in the middle, having to wrestle with using cheaper gaming cards with features that don't quite match up with their needs.

Nvidia needs to start making lower-end Quadros with better specs for 3D freelancers (a sort of new line which is aimed specifically at our needs).

Then it would be clear which demographic should use which kind of card, and we would all be happy (gamers, freelancers, studios).

Share this post


Link to post
Share on other sites

I think Nvidia takes features out of the gaming cards in places where they believe the gamers won't notice.

Unfortunately there are a bunch of freelance 3D artists who use these gaming cards for digital content creation, and I have a feeling those artists are in the minority.

The big 3D studios still use render farms which are mostly CPU based. If the big studios still think they might possibly need more GPU power, then they will buy the top end Quadro cards.

That leaves the freelance 3D artists stuck in the middle, having to wrestle with using cheaper gaming cards with features that don't quite match up with their needs.

Nvidia needs to start making lower-end Quadros with better specs for 3D freelancers (a sort of new line which is aimed specifically at our needs).

Then it would be clear which demographic should use which kind of card, and we would all be happy (gamers, freelancers, studios).

What they need to do is bring Quadro cards down to the realm of their high end consumer cards. There is no difference hardware-wise, and software-wise, they aren't spending any more on quadro drivers than GeForce drivers. It's price-gouging...plain and simple. Small to medium studios aren't interested in getting gouged either.

Share this post


Link to post
Share on other sites

How can your views (with which I agree) be brought to the attention of Nvidea. They would sell a lot more GFX cards than Quadros, It's similar to lowering taxes to gain more revenue.

Share this post


Link to post
Share on other sites

quadro cards are useless for gaming.. :(

 

makes me sad.

 

I think if they just released CUDA cards people would buy them and it would bring in more profit for them... kinda compromise.

 

but yeah its just price gouging.. they dont want pro's using cheap gaming cards.

Share this post


Link to post
Share on other sites

What bothers me about this price gouging is that they know we know they are doing it, and continue anyway. It doesn't take a rocket scientist to determine that the hardware, itself is no different (except maybe a bit more RAM in some models), and that NVidia is simply charging 300-500% more for just different drivers. Absurd. But Nvidia is not alone in this. AMD does the same thing with their Pro Cards.

 

Nvidia knows that much of the CG industry...not just freelance artists...rely on the consumer line, rather than the price-gouging workstation cards. That's why CUDA and PhysX is on the consumer line, too. All the GPU rendering vendors typically use the GeForce models in their demonstrations. The Titan seems more like a workstation card sold as a Geforce model. So, at least it's there.

 

Another encouraging thing is that the gaming market is really pushing things forward, and forcing NVidia to beef up RAM capacity, with all the 4K monitors coming out now. After Maxwell, Volta is supposed to offer stacked memory and a much larger memory bus. So, I think I'll just grab one of these 780 6GB, when I can, and that should hold me over until that Volta line comes out. It's amazing, though, that this GTX 580 is still kicking tail 3yrs later.

Share this post


Link to post
Share on other sites

I'm not 100% sure but I have heard that there is a difference between Quatro's and the gaming cards.  The Quatro's come with ECC memory which auto corrects memory errors and they have better support for floating point operations in the driver.  I think the biggest difference is in the support that you get with Quatro's in that you can call up Nvidia and they will fix whatever problem you are having, one way or another.  I don't know if that justifies the price difference or not though.  <_<  Also I have heard that the Quatro driver has optimized OpenGL code and so they can handle higher number of polys in the view port than the gaming cards.

 

If you have the money, the best system (for GPGPU) is a Quatro card for the display and a pack of high-end gaming cards for computing.  That way you get the best of both worlds.   :)

Edited by Grimm

Share this post


Link to post
Share on other sites

damn that strix 780 looks awesome.

 

too bad i cant find an cuda benchmarks for it..

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×