Anyone Know When The Next Gen Nvidia Cards Are Being Released

[kalniel]

Pricing for any of these? Well supposed RRP.

We don't even know what models there will be yet.

[CAT-THE-FIFTH]

Supposedly,the GTX680 will be launched on the 22ND and it will also only use two PCI-E 6 pin power connectors:

http://www.overclock.net/t/1225986/t...#post_16691060


http://translate.google.fr/translate...%23post9205955

The price is supposed to be around $550.

[kalniel]

Supposedly,the GTX680 will be launched on the 22ND and it will also only use two PCI-E 6 pin power connectors:

http://www.overclock.net/t/1225986/t...#post_16691060


http://translate.google.fr/translate...%23post9205955

The price is supposed to be around $550.

That's more like it. Scan are showing it off in an event on the 24th, which is a Saturday, so a release a few days before it quite likely. And the price is more reasonable.

[Domestic_Ginger]

Price is more as expected? Reasonable does not seem to be.

[kalniel]

Price is more as expected? Reasonable does not seem to be.

Poor choice of words on my part, agreed. The rumour is more reasonable.

Basically it looks like it's going to be about the same performance as a 7970 - I expect wins in PhysX titles, and perhaps losses in shader heavy titles and at higher resolutions. So they're pricing it the same as a 7970. If they produce enough parts they might release a version to compete with the 7950 but for now that's looking unlikely.

That means the only competition nVidia have for the 7950 is the 580, which is inferior in nearly all titles, hence the price cut. IF AMD have enough cards, they could hurt nVidia here by keeping the 7950 agressively priced, though they've got to watch the toes of the 7870 I guess.

[CAT-THE-FIFTH]

Some more pictures of the GTX680:

http://www.techpowerup.com/162272/Ne...t-The-Web.html

A dual GK104 card is in the works:

http://www.techpowerup.com/162275/Du...es-in-May.html

The $550 price is also mentioned here and the GK104 is meant to have better performance per watt than the HD7900 series:

http://www.fudzilla.com/home/item/26...gainst-hd-7970

Kepler might end up being quite efficient.

Edit!!

Some details of the GPU:

http://www.chiphell.com/forum.php?mo...=2#pid11424899

It is meant to have 3.54 billion transistors. According to the person who leaked the latest pictures they have no working drivers so cannot bench the cards.

The main reviewer at HC did say this though:

http://www.xtremesystems.org/forums/...87#post5067987

[CAT-THE-FIFTH]

The GTX680 has a 1GHZ+ core clockspeed:
http://forum.beyond3d.com/showpost.p...postcount=2489

Nvidia performance chart showing upto 40% speed increase over HD7970:
http://www.techpowerup.com/162305/Ge...70-NVIDIA.html

[Hicks12]

But will Nvidia have fixed their microstutter type issue? I doubt it... which makes average FPS reviews a bit useless.

I think its pretty obvious that Nvidia is going to trump AMD, after all this is AMDs true change in design, it would be pretty hard for Nvidia to fail at beating an architecture in its infancy (*cough*(Like Fermi*cough*). AMD will however have the extreme upper hand when it comes to next year as you would think they could change the design alot more and grab performance from pretty much everywhere.

Nvidia will take the crown, power effciency better than AMD though is a bit of an odd one as Nvidia are always the big ass guys that pump out high fps for a relative wattage.


Still interested to see what Nvidia really have been doing after fermi though

[Terbinator]

But will Nvidia have fixed their microstutter type issue? I doubt it... which makes average FPS reviews a bit useless.

I think its pretty obvious that Nvidia is going to trump AMD, after all this is AMDs true change in design, it would be pretty hard for Nvidia to fail at beating an architecture in its infancy (*cough*(Like Fermi*cough*). AMD will however have the extreme upper hand when it comes to next year as you would think they could change the design alot more and grab performance from pretty much everywhere.

Nvidia will take the crown, power effciency better than AMD though is a bit of an odd one as Nvidia are always the big ass guys that pump out high fps for a relative wattage.


Still interested to see what Nvidia really have been doing after fermi though

Even with a new architecture AMD's cards should have been better. Nvidia were the pipe cleaner in this regards so AMD should have known where resources were best allocated and in addition, games have been pumped out with the 'Compute' architecture being the new standard to address.

As always with these swings and roundabouts scenarios, on the surface its hard to see Nvidia not taking the next round as well given the tracks laid by Fermi/Kepler.

[Hicks12]

Oh yes they should have been better, it would be worse i think if Nvidia hadnt been the "trial" company , they took the risk with fermi and have made AMD thing seriously about the new architecture. I still see AMD being able to bring out a ton of performance from driver updates but in the end its still down to the fact we have nothing to take advantage of these technologies, crysis is the only game that can really get intense but its getting old and so doesnt take advantage of this architecture .

Hopefully we get some excellent games this year or the next to make use of the cards AMD and Nvidia have made, BF4 (if its made?) will certainly help push it .

And yes that is pretty much the microprocessor industry defined, swings and roundabouts . Nvidia likely this year, AMD next year and then maybe Nvidia again. Im hoping that AMD can finally get back on that roundabout and get to be top dog in the CPU market for a generation or two, would really help them.

[CAT-THE-FIFTH]

the surface its hard to see Nvidia not taking the next round as well given the tracks laid by Fermi/Kepler.

I would be tad cautious about making predictions for more than one generation though,ie,the current one. Anyway, I would wait until actual reviews are out as the leaked chart indicates something I suspected. It will be interesting to see if reviews show the same thing though.

[CAT-THE-FIFTH]

Actually look at the table in posts 153 and 154 - it seems the chart might be based on it!

[CAT-THE-FIFTH]

PC Perspective say they will have a Kepler preview up soon:

https://twitter.com/#!/ryanshrout/st...08075253567488

It looks like a preview of the GT640M:

http://www.pcper.com/reviews/Mobile/...Arrives-Mobile

The GT640M has 384 Kepler cores running at 625MHZ and has the shaders running at 1250mhz with a 128 bit memory controller and 900MHZ DDR3 . The GT555M used in the review has 144 Fermi cores running at 590MHZ with a 192 bit memory controller using DDR3. Despite lower memory bandwidth the GT640M was running BF3 around 15% faster than the GT555M! Just Cause 2 was around the same speed on the GT640M too.

Considering the GT555M was running in a full sized Alienware M14X laptop and the Acer was using a low voltage CPU,it looks a decent result.

It also seems Kepler still has hotclocks too.

Edit!!

It seems due to the Turbo Boost technology,the GPU was running quite hot(it seems Acer set the boost margins relatively high).

[Hicks12]

Is it just me cat or is that result pretty bad? I that's making it seem like me kpler cores are worth about half of Fermi cores.. So amd have little to worry about

[kalniel]

Nvidia performance chart showing upto 40% speed increase over HD7970:
http://www.techpowerup.com/162305/Ge...70-NVIDIA.html

That's a chart from ages old 'leaked' data. Nothing new.

[Domestic_Ginger]

The efficiency card is one ATI have been forced to play in the past; graciously conceding the performance crown.

I'm thinking if this card did anything special we would know about it by now ala fermi's tessellation (enough s's and l's and it should be spelt okay); the large man would have come out and told us about it.