Geforce Titan - GK110 will be a consumer part

[Andrew LB]

There is/was no GK100, at least not in any conceivable form. Drawing boards maybe.

nVidia has always used the code name GK110 which actually had it's first tape out in December 2011 and was completely functional in May, 2012.

[Handrox]

DUAL GK110, will be possible?

[kromosto]

DUAL GK110, will be possible?

single one with this price ... maybe ... at most 10 of them just for show business.

[Tha Last Meal]

DUAL GK110, will be possible?

I do not think so, but may Dual GPU come from the GK114 -not GK110......Who knows

[Handrox]

Imagine a sigle GPU POWER GK110 accompanied by a DUAL POWER GK110 is something very interesting.

[Tha Last Meal]

Imagine a sigle GPU POWER GK110 accompanied by a DUAL POWER GK110 is something very interesting.

Certainly you will be able to cook it

[vitchilo]

Dual GK110 on a single GPU with a waterblock + radiator + pump included would be quite nice. All for 500$... please?

[Handrox]

Certainly you will be able to cook it

Is it? nothing that does not resolve a WC

[PedantOne]

ARES II is Water Cooled too, why not dual GK110

[vitchilo]

ARES II is Water Cooled too, why not dual GK110

At the price they are selling it (1500$) they should do Dual GK110... otherwise it's way overpriced.

[Tha Last Meal]

Is it? nothing that does not resolve a WC

if we're talking about the heat of Dual GK110 There other solutions such waterblock And the rest of the pieces their own.this may Happen....but i don't think so.
But if that happened.....damn that will be a big Bodybuilder muscles dogg

[Handrox]

1000€...


Link

[Tha Last Meal]

1000€...


Link

GTX Titan will come 800$-900$ I'm sure of it

[kromosto]

imo it is the entry price in 2 weeks it will settle

[Handrox]

http://www.2compute.net/Asus/GTXTITAN-6GD5.htm

if we're talking about the heat of Dual GK110 There other solutions such waterblock And the rest of the pieces their own.this may Happen....but i don't think so.
But if that happened.....damn that will be a big Bodybuilder muscles dogg

that price makes me very discouraged, if maybe lower the price of the GTX 680 I make an SLI ...

[Tha Last Meal]

that price makes me very discouraged, if maybe lower the price of the GTX 680 I make an SLI ...

the price of GTX 680 will be lower I think between $400-$440, soon

you can delete this post by go to Edit Post and than to Delete Post and Delete this post in the following manner

[Lanek]

Waterblocks, allready ready for Titan .

Got a mail from EK waterblock.. a poll for vote for the design of their next waterblock.. for the Titan..

EK Water Blocks, Ljubljana based premium water cooling gear manufacturer, has opened a poll on EK-ThinkCell site where you can vote and choose your favourite design for the upcoming nVidia Geforce Titan Full Cover series water block as well as the redesign of the existing EK-Supremacy CPU water block.

http://thinkcell.ekwb.com/idea/new-f...oose-your-best

Just for the record. The VGA card in the back is NOT Geforce Titan!

( all my hope of getting a pcb screen ruined )

[metachronos]

Waterblocks, allready ready for Titan .

Got a mail from EK waterblock.. a poll for vote for the design of their next waterblock.. for the Titan..



http://thinkcell.ekwb.com/idea/new-f...oose-your-best


( all my hope of getting a pcb screen ruined )

Hopefully the clean design without all those circles wins.

[Stevethegreat]

nVidia has always used the code name GK110 which actually had it's first tape out in December 2011 and was completely functional in May, 2012.

And the question is : why -then- nvidia hasn't released it around that time to completely dominate the market given that AMD would had seemed to be unable to respond for -close to-two years in a row.

Is there a weird law against monopoly which would disable one of the competitors to release something that far better from competition? It seems senseless to me that nvidia gave the chance to AMD to breathe and even momentarily take the crown (driver update), what was the point of that?

[hennyo]

I normally don't post but it really seems to me the vast majority of people here and around the net are overlooking the most basic detail with this card.

When a company has a product, one of of dividing into two different categories is, products that are made to sell, and products that are sold because they are made. This may seem like the same thing, but it is very different, almost all gpu's made for consumers are the first category, while Titan is a rare exception and falls into the second category.

Titan gpu's are not made to be sold as Titan's, they are K20x's that don't quite make it. So what they do is take a k20x, disable one of the core clusters from 15 to 14, remove the ECC memory, and then use the bios to turn DP units into integer unit to increase the graphics performance while killing their server use in one blow.

So when your asking yourself why didn't these come out earlier, it was a function of how long it took them to get a stockpile of these mostly working but not quite par chips. Also the sale price is the same, to get as much money from them they can, without gaining too much ire from consumers for exploiting the demand.

So there it is, this gpu is not made due to any sort of consumer demand verse profit made, so stop thinking in terms as if it was. The single gpu performance crown and all that other stuff here, is just bonus, nothing more.

[boxleitnerb]

K20X has 14 clusters. And it is not really clear if the DP units are SP capable. Assuming there are dedicated DP units, that is. That much is not really clear.

[Lanek]

K20X has 14 clusters. And it is not really clear if the DP units are SP capable. Assuming there are dedicated DP units, that is. That much is not really clear.

I was ask me the same thing some days ago..

Anandtech presentation of K20x

Fundamentally GK110 is a highly enhanced if not equally specialized version of the Kepler architecture. The SMX, first introduced with GK104, is the basis of GK110. Each GK104 SMX contained 192 FP32 CUDA cores, 8 FP64 CUDA cores, 256KB of register file space, 64KB of L1 cache, 48KB of uniform cache. In turn it was fed by 4 warp schedulers, each with two dispatch units, allowing GK104 to issue instructions from warps in a superscalar manner.

GK110 builds on that by keeping the same general design, but tweaking it for GK110’s compute-focused needs. The single biggest change here is that rather than 8 FP64 CUDA cores GK110 has 64 FP64 CUDA cores, giving it 8 times the FP64 performance of a GK104 SMX. The SMXes are otherwise very similar at a high level, featuring the same 256KB of register file space, 64KB of L1 cache, 48KB of uniform cache, and the same warp scheduler structure. This of course does not include a number of low level changes that further set apart GK104 and GK110.

[hennyo]

Opps delete post I was mistaken.

K20x while it has 15 clusters, only 14 are active, K20 has 13 like I thought.

Details from the anandtech review of the K20, http://www.anandtech.com/show/6446/n...rrives-at-last .

[BenchZowner]

And the question is : why -then- nvidia hasn't released it around that time to completely dominate the market given that AMD would had seemed to be unable to respond for -close to-two years in a row.

Is there a weird law against monopoly which would disable one of the competitors to release something that far better from competition? It seems senseless to me that nvidia gave the chance to AMD to breathe and even momentarily take the crown (driver update), what was the point of that?

Leading another company to bankruptcy in a 2 player battle ends the battle, and the winner doesn't really win anything ( anti-monopoly laws mainly )

[3lfk1ng]

...AMD would had seemed to be unable to respond for -close to-two years in a row.

Due to driver enhancements last year (specifically 12.11), AMD managed to outperform the GTX680 with the HD7970s by the end of 2012(GCN latency issues aside) and costs a full 20% less. (Link)