Kepler Nvidia GeForce GTX 780

[GoldenTiger]

Thanks Olivon! Looks like there may be a large fan on these things?

[Kuroimaho]

PHK has posted informations regarding GTX 680, 6+8 pins or 2X6 pins (670), 225W TDP :

That's very odd, could it be an ES board ?

I love everything what makes these card run more silent, so I really appreciate experiments to improve cooling.

[HuffPCair]

Dang looks like those links no longer work.

[GoldenTiger]

Dang looks like those links no longer work.

Yeah, most of the stuff was edited out by the time I had even posted a reply earlier .

[marten_larsson]

it is true that amd's refresh (8000 series?) wont be too far off when nvidia releases kepler but it will simply be that, a refresh. the 7000 series is a brand new architecture, kepler is a brand new architecture. since they are both brand new architectures and not refreshes i do agree people should wait to see who has a better leg to stand on.

How is AMD's not even announced 8000-series known to be closer to a refresh of GCN than Kepler is to Fermi? I thought Fermi was the new architecture and what we'll see from Nvidia will be upgrades of that arch but nothing revolutionary like GT200/300 - GF100.

[SKYMTL]

How is AMD's not even announced 8000-series known to be closer to a refresh of GCN than Kepler is to Fermi? I thought Fermi was the new architecture and what we'll see from Nvidia will be upgrades of that arch but nothing revolutionary like GT200/300 - GF100.

The post you replied to makes me really wonder how many posters are actually plants by marketing teams trying to divert attention from a competitors' product. But then again, there will always be someone who is "disappointed" or "waiting".

The fact of the matter is that AMD's next generation parts are nowhere near release. Think more along the lines of 2013 considering Tahiti refreshes haven't even taped out yet, let alone a move towards a more advanced architecture.

In addition, I think people have to be clear about one thing and I've been saying this from the start: The first generation of high end GCN parts were never meant to compete against Kepler. They were originally slated to be fabricated at 32nm and be a a direct competitor to NVIDIA's refreshed Fermi cards (GTX 580, GTX 570, etc.). Their performance results pretty much back this up as well.

IF (and that is a big IF) NVIDIA has designed Kepler to be a real step up from Fermi then AMD's current generation doesn't stand a chance in the high end market. On the other hand, AMD has proven that they can compete on price and all of their HD 7000-series (particularly in the $200+ price brackets) have more than enough padding to endure cuts of 10% or slightly higher.

[cold2010]

http://translate.google.com/translat...%2F%3Fp%3D5264

[highoctane]

http://translate.google.com/translat...%2F%3Fp%3D5264

Cool, a real photo at least makes it feel like things are getting closer.

[SKYMTL]

That's what we call a test mule. It is an engineering card with localized PCB I/O connectors for diagnostic equipment.

[tbone8ty]

still have a long ways to go for the announcement of the announcement

[N19h7m4r3]

still have a long ways to go for the announcement of the announcement

Are you ready? To be ready?!

Hah, that Fermi announcement was a laugh.

[SKYMTL]

I'll go out on a limb here and predict you won't see any "announcements of announcements" this time around.

[Manicdan]

can people measure the chip size from that last photo? it actually looks normal sized (like 300-350mm2)

[Olivon]

[/IMG]http://www.arabpcworld.com/images/2012/03/nvidia-GK104-.jpg[/IMG]

http://translate.google.com/translat...%2F%3Fp%3D5264

Original link : http://www.chiphell.com/forum.php?mo...8&pid=11248448

[GoldenTiger]

From h...

Old Today, 12:15 PM
Kyle_Bennett HardOCP Editor-in-Chief, 14.8 Years

I am seeing information out of China this morning showing 45% to 50% performance increase over 580 in canned benchmarks.

Not seeing anything on the "usual suspects", that would be pretty nice.

[tajoh111]

If this is being called a gtx 680, this doesn't bode too well for pricing.

[bingo13]

I'll go out on a limb here and predict you won't see any "announcements of announcements" this time around.

I have no idea what is worse, all the assumptions made in this thread or the one in the IVB thread.

[aussie-revhead]

If 660 is looking at being released fairly soon then how long would they hold back 680? Until AMD catch up? Maybe that's why a long delay has been hinted until 680.... That sucks

[K404]

I was under the impression that the first release wouldn't take back the crown and the card that WOULD.... is delayed.

[DilTech]

I was under the impression that the first release wouldn't take back the crown and the card that WOULD.... is delayed.

This card should actually do the job... NVidia put out a lot of misinformation to find the leaks once again, which is where it not beating the 7970 apparently came from.

[highoctane]

This card should actually do the job... NVidia put out a lot of misinformation to find the leaks once again, which is where it not beating the 7970 apparently came from.

I really don't know which rumors to believe but personally I feel its better to have mediocre expectations and be surprised rather than big hype and big let down.

One set of rumors puts it around 580 level and another set of rumors has it unbeatable by anything currently on the market.

[DilTech]

I really don't know which rumors to believe but personally I feel its better to have mediocre expectations and be surprised rather than big hype and big let down.

One set of rumors puts it around 580 level and another set of rumors has it unbeatable by anything currently on the market.

I play it by common sense and some well educated guessing.

If the shader count is true, even without the hot-clocks, there's just no way this thing is only going to perform at 580 levels. Even though it only has a 256-bit memory bus, the ram speeds are apparently a ghz higher than what we saw in the GTX 580 (albeit it would need a memory clock of 6 ghz to match the bandwidth of the 580)...

Now, I know what people are thinking... "with that big of a memory bandwidth handicap there's no way it can beat the 7970, and will only be around 580 performance"... I'll happily remind people of the fact that memory bandwidth is only part of the story. The 8800GTX had a 384-bit memory bus, while the 8800GTS 640mb had a 320bit, we all remember that. Meanwhile, the refresh for the 8800GTS and 8800GTX were both 256bit and slightly higher memory clocks (NOT enough to close that gap in bandwidth). How were they able to stand above in most cases (the 8800GTX beat the 9800GTX when the 9800GTX hit it's memory barrier) with lower memory bandwidth? Efficiency. One thing that was heavily tweaked for the G92 was it's efficiency in usage of its bandwidth. Now mind you, the 9800GTX only had a 60'ish mhz clock speed increase while losing more than a few parts and STILL beat the 8800GTX for the most part.

Fast forward to now, you're talking about a very different situation. 3x more shaders and 33% more rops (and who knows what else) and a bit of a memory bandwidth reduction. You'd have to be crazy to expect this card to only tie the 580. I'd say ~50% faster than the 580 sounds about right. That's not a crazy talk assumption, that's going by NVidia's own history and what we know so far.

Anyone expecting 100% across the board though, THAT is a pipe dream if I've ever heard of one.

[kuroikenshi]

Anyone expecting 100% across the board though, THAT is a pipe dream if I've ever heard of one.

LOL I would not ditch pipe dreams since most of the fan-base assumptions on many brands "and forum threads" thrive on it


I like your example and since I have experienced it first hand I can also concur with your statement.

[highoctane]

I play it by common sense and some well educated guessing.

To me common sense in this type of situation has shown that new uarch plus late launch/delays mixed with possibly poor yields should be cause enough to pause and think about the situation. We've all seen this situation now numerous times over the years, everyone here reading these threads should be feeling a little dejavu.

We've seen plenty of hardware that looks killer in the specs and considered the second coming in community hype only to be so so in reality, doesn't matter the brand.

Not that I doubt kepler can or will be a great product, simply kepler has yet to prove itself to me to be a great product.

[DilTech]

To me common sense in this type of situation has shown that new uarch plus late launch/delays mixed with possibly poor yields should be cause enough to pause and think about the situation. We've all seen this situation now numerous times over the years, everyone here reading these threads should be feeling a little dejavu.

We've seen plenty of hardware that looks killer in the specs and considered the second coming in community hype only to be so so in reality, doesn't matter the brand.

Not that I doubt kepler can or will be a great product, simply kepler has yet to prove itself to me to be a great product.

Late really doesn't mean as much as we make it out to be. While to us 3 months seems like a huge gap, in the grand scheme of things it's truly not. In most markets, there is usually a gap between competing products. Video game consoles, for example. The PS2 launched a year before the xbox, the dreamcast a year before the ps2. The xbox360 launched a full year before the PS3... 3'ish months isn't late, considering it takes longer than that for a new graphics series to even make an indication of its own existence in their market share or sales--especially when there isn't a game out right now that can't manage just fine on last gens hardware. It just annoys the enthusiasts to have to wait. Personally, I was hoping the card would've launched by my b-day, since my woman was going to buy me one for my bday. I told her to give me a gift card for the money amount to microcenter. I'll wait and see for myself who has the better card.

Fact is, a rather sizable majority of the market don't do their research. They don't know a company has a new product out, or they have a new product coming out for that matter. They walk into their local store (which is likely employing people who also don't know anything), tell the guy in the blue shirt that they have a said amount of money to spend, and the guy in the blue shirt says "buy this". They then pay someone in a white shirt at the same store $50+ to put the card in for them because they don't know how to do that, and go on their merry little way home. I can honestly say 3 out of 5 people who's systems I work on are still running the drivers that came on the disc with their cards, since they haven't updated since the initial installation! Considering how many systems I've worked on, that pretty bad odds.

As much as people like to claim doom and gloom every time a company is behind the opposition's release cycle, the market really doesn't care, because the market doesn't even know. To us (who mind you, are NOT the majority of the market...enthusiasts are always the last person noticed by big companies), what should matter is who has the best product for the amount of money we are willing to spend. To them, what matters is what they were told to buy by someone else.

It's just like SkyMTL said... The 7970 was suppose to be the 6970 before TSMC cancelled their 32nm process. AMD just made a few changes to the design and released it as the 7970. So in that sense NVidia isn't late, AMD was just early. Now Fermi, on the other hand, was late.