• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

First NVIDIA GeForce Titan 780 Performance Numbers Revealed

Lol, yes. But really, it is entertaining to see people get carried away with obviously bad/fake information. :)

I think TPU should have an official turd pic for instances just like this...Perhaps one of somebody holding a fresh squeezed one in their hand with a toilet in the background...just saying
 
oh my freaking graphics card. ima get one
 
If this is true I am excite!
 
Hold up a freaking second...this directly contradicts the other rumour

Other reports suggest that the upcoming GeForce Titan will also not be called the GeForce GTX 780, nor will it take a GTX 600 series name. It will simply be called the GeForce Titan, likely inspired by the by the Cray Titan supercomputer, which houses nearly 19 thousand Nvidia Tesla K20X cards.

http://www.tomshardware.com/news/GeForce-Titan-GK110-GTX690,20797.html

I can see the GTX branding on the screenshot. And to add insult to injury, this moronic Chinese site called it a GTX 780 Ti. Class A bullshit right there. Even if the Titan had all 2880 cores/15 SMX units enabled, it cannot possibly beat a GTX 690 on paper (or carry a mid-range series name). The only way it could even come close would have to be at some ridiculously VRAM-intensive resolution, like 3840x2160 or higher and even then it would be a struggle. Don't believe a word of this shit.

Oh and it won't have 6GB of VRAM. Not a chance in hell of that happening. If we take a look at Nvidia's timeline (reference designs only, 3rd party non-reference designs with extra VRAM are NOT included in this comparison)

7000 to 8000 series (512MB to 768MB) = 50% more VRAM
8000 to 200 series (768MB to 1024MB) = 33% more VRAM
200 to 400 series (1024MB to 1536MB) = 50% more VRAM
400 to 600 series (1536MB to 2048MB) = 33% more VRAM

as you can see, they follow the same pattern. Since the 700 series is essentially a resale of the same architecture, the best you can expect is 3GB of VRAM. Since this is their king of the hill halo card just like their GTX 690, you will not see 3rd party re-designs with double the VRAM since Nvidia will not allow it. Oh and since Nvidia's partners will still have loads of GTX 690s to get rid of, and the Titan will be priced under the 690, therefore the Titan will never beat the GTX 690, at least not in gaming.

This whole story reeks of horse shit. 85% of performance of a GTX 690 sounds about right, and that's the best anyone should expect.
 
The only way it could even come close would have to be at some ridiculously VRAM-intensive resolution, like 3840x2160 or higher and even then it would be a struggle.

Anyone that buys a top of the line GPU to play anything less than high resolution is--generally speaking--a fool. Nobody buys a GTX690 to play WoW at 800x600.

This whole story reeks of horse shit. 85% of performance of a GTX 690 sounds about right, and that's the best anyone should expect.

For starters, it's a rumor. The first line of the post reads "The rumor mill is spinning". If you're taking this as fact you're doing it wrong. Also, even 85% the performance of a GTX690 would be pretty damn impressive.
 
It's somewhat very unlikely that performance will be more than 2x of its predecessor. Usually 2x is the limit at which they hold onto. Though they rarely do it. For AMD i only remember it happening with HD4800 -> HD5800 transition where HD5800 was twice as fast.

3870 -> 4870 was 2.5 times performance increase!
 
3870 -> 4870 was 2.5 times performance increase!

That's more because of the huge issues ATi had at the time of producing that card. Was a huge mis-step on their end.

Same with the 79xx series cards in a way. Although all we have is speculation as to why nVidia chose to release GK104 as their top end chip rather than GK100.
 
That's more because of the huge issues ATi had at the time of producing that card. Was a huge mis-step on their end.

Same with the 79xx series cards in a way. Although all we have is speculation as to why nVidia chose to release GK104 as their top end chip rather than GK100.

Whatever the reason is, the fact is that they couldn't do it.
 
Whatever the reason is, the fact is that they couldn't do it.

Couldn't, or wouldn't? We just don't know. No information was released, and nVidia released their usually mid-range GPU as a top tier product. Mainly because AMD's performance just wasn't there.
 
Couldn't, or wouldn't? We just don't know. No information was released, and nVidia released their usually mid-range GPU as a top tier product. Mainly because AMD's performance just wasn't there.

...but it's not like the GTX 680 walked over the 7970. If they had something faster that was working would have made those number easily favor nVidia if that is the case. You're right though, we don't know, so something faster than the 680 could have been ready, then again it might not. It also did take a while for nVidia to roll out the 680, so I think to say a faster card was ready to be marketted would be a hell of a stretch.
 
...but it's not like the GTX 680 walked over the 7970. If they had something faster that was working would have made those number easily favor nVidia if that is the case. You're right though, we don't know, so something faster than the 680 could have been ready, then again it might not. It also did take a while for nVidia to roll out the 680, so I think to say a faster card was ready to be marketted would be a hell of a stretch.

It's true the 680 didn't walk all over the 7970, but nVidia realized they could match the performance of the 7970 with their GK104 chip, and decided to roll that out instead of the GK100. Whether they were able to produce great quantities of the GK100 or not we will never really know. However, if the 7970 performed at the level that the GK100 was supposed to be capable of, you bet your bacon that nVidia would have held off longer and released that chip as top tier.

Take a look at what they did for GF100. nVidia waited a LONG time to release their GF100 because they had yield issues and needed something that would match the 5870.

EDIT: Additionally, nVidia typically releases their top tier first then back fills the line up. So it's entirely possible GK100 was ready to go, and they decided to wait until GK104 was ready and released that to compete with the 7970.
 
...but it's not like the GTX 680 walked over the 7970. If they had something faster that was working would have made those number easily favor nVidia if that is the case. You're right though, we don't know, so something faster than the 680 could have been ready, then again it might not. It also did take a while for nVidia to roll out the 680, so I think to say a faster card was ready to be marketted would be a hell of a stretch.

They were in the position where they could release a mid-range card, as a high-end card (according to the gaming performance of their competitors) and double their profit margins. They would have been insane not to do it--from a business perspective. If AMD had released the HD7970, and it was 15-20% faster than it was at launch, Nvidia would have launched GK104 as the GTX660, GTX650Ti, and the GTX650, and filled the $200-300 price range. Instead, they were able to overclock it to high hell and sell it for $350-550. Why would they not do that? It meant for every GTX680 they would sell, they'd make something like $200 in pure profit.
 
Back
Top