Wednesday, March 7th 2012
![NVIDIA](https://tpucdn.com/images/news/nvidia-v1723816949774.png)
GeForce GTX 680 Features Speed Boost, Arrives This Month, etc., etc.
Here are some key bits of information concerning the upcoming GeForce GTX 680, a performance single-GPU graphics card based on the 28 nm GK104 GPU by NVIDIA. The information, at face value, is credible, because we're hearing that a large contingent of the media that finds interest in the GPU industry, is attending the Game Developers Conference, where it could interact with NVIDIA, on the sidelines. The source, however, is citing people it spoke to at CeBIT.
First, and most interesting: with some models of the GeForce 600, NVIDIA will introduce a load-based clock speed-boost feature (think: Intel Turbo Boost), which steps up clock speeds of the graphics card when subjected to heavy loads. If there's a particularly stressing 3D scene for the GPU to render, it overclocks itself, and sees the scene through. This ensures higher minimum and average frame-rates.
Second, you probably already know this, but GK104 does indeed feature 1,536 CUDA cores, which lend it a strong number-crunching muscle that helps with shading, post-processing, and GPGPU.
Third, the many-fold increase in CUDA cores doesn't necessarily amount to a linear increase in performance, when compared to the previous generation. The GeForce GTX 680 is about 10% faster than Radeon HD 7970, in Battlefield 3. In the same comparison, the GTX 680 is slower than HD 7970 at 3DMark 11.
Fourth, the NVIDIA GeForce GTX 680 will very much launch in this month. It won't exactly be a paper-launch, small quantities will be available for purchase, and only through select AIC partners. Quantities will pick up in later months.
Fifth, there's talk of GK107, a mid-range GPU based on the Kepler architecture, being launched in April.
Next up, NVIDIA is preparing a dual-GPU graphics card based on the GK104, it is slated for May, NVIDIA will use Graphics Technology Conference (GTC), as its launch-pad.
Lastly, GK110, the crown-jewel of the Kepler GPU family, will feature as many as 2,304 CUDA cores. There's absolutely no word on its whereabouts. The fact that NVIDIA is working on a dual-GK104 graphics card indicates that we won't see this chip very soon.
Source:
Heise.de
First, and most interesting: with some models of the GeForce 600, NVIDIA will introduce a load-based clock speed-boost feature (think: Intel Turbo Boost), which steps up clock speeds of the graphics card when subjected to heavy loads. If there's a particularly stressing 3D scene for the GPU to render, it overclocks itself, and sees the scene through. This ensures higher minimum and average frame-rates.
Second, you probably already know this, but GK104 does indeed feature 1,536 CUDA cores, which lend it a strong number-crunching muscle that helps with shading, post-processing, and GPGPU.
Third, the many-fold increase in CUDA cores doesn't necessarily amount to a linear increase in performance, when compared to the previous generation. The GeForce GTX 680 is about 10% faster than Radeon HD 7970, in Battlefield 3. In the same comparison, the GTX 680 is slower than HD 7970 at 3DMark 11.
Fourth, the NVIDIA GeForce GTX 680 will very much launch in this month. It won't exactly be a paper-launch, small quantities will be available for purchase, and only through select AIC partners. Quantities will pick up in later months.
Fifth, there's talk of GK107, a mid-range GPU based on the Kepler architecture, being launched in April.
Next up, NVIDIA is preparing a dual-GPU graphics card based on the GK104, it is slated for May, NVIDIA will use Graphics Technology Conference (GTC), as its launch-pad.
Lastly, GK110, the crown-jewel of the Kepler GPU family, will feature as many as 2,304 CUDA cores. There's absolutely no word on its whereabouts. The fact that NVIDIA is working on a dual-GK104 graphics card indicates that we won't see this chip very soon.
105 Comments on GeForce GTX 680 Features Speed Boost, Arrives This Month, etc., etc.
Basically they are launching a mid tier card under a top tier number with no announcement. Its BS if true.
which makes me wonder what are they going to call the top end gk110 cards on Q4?
A name is just a name. FYI I was monstly being sarcastic, not that I'm not right on that. Nvidia has always been adjusting to AMD's naming change BS, if you want to call it that way, first with HD2000 to HD3800, then with HD5000 to HD6000 and last (maybe) with HDx800 becoming HDx9xx so they call it GTX680, so what?
And I'm not saying I like all the games that both AMD and Nvidia play, but complaining about this particular case with the things they've done in the past is rather stupid.
I bet they charge 580 prices for 560 performance per the 6xx generation so to speak. Its underhanded. People who dont know chip numbers will think they are getting a top tier card unless NVIDIA announces it. They will be paying for a top tier card and getting a mid-range. Its shady as hell and you know it.
I don't like it anymore than you do, but I'm just being realistic. If it beats Tahiti, why shouldn't they call it GTX680 and ask a price accordingly. I remind you that YOU are the defendant of capitalism, remember? I'm the socialist (not me, my country). ;)
I have mine humming along at 1,200 core so far :) It's not mid-range b/c it's the fastest card they'll have till June or who knows when.
AMD can just bump clocks 2-300 mhz and b**** slap nvidia (not sure if they will since they like power savings, but I think they should).
Plus, they can cut prices some if necessary with improving yields. Nvidia won't have that for many months lol
and its allegedly better then an un clocked 7970 in compute only , we all know that balls to the wall the 7970 will clock higher and ultimately perform better then this speculation would OC imho, as they are already clearly capable of higher clocks Ala the boost feature, just not capable of maintaining them over time as i see it , it dosnt sound like OCérs delight:confused:
We need that nvidia GTX600 make a kickass GPU!, to see the price going down in in the two sides. :shadedshu
If that's the case we're pretty screwed.
Just as well AMD doesn't start ripping off Intel's naming conventions for its APU's wholesale -including the "K" unlocked multi suffix...oh, wait...
I guess on company's "sand bagging" is another company's astute business acumen
(EDIT: A quick browse of TPU's back articles on AMD's GPU/APU naming conventions doesn't seem to yield any concerns from you. Is this practice something that has recently earned your ire?) Getting your panties in a wad? Over what?...an "if" ?
Who says "GK104" is a midrange GPU ? all the scuttlebutt is tending towards the same ballpark die space, compute perf, overall perf as Tahiti. You see Tahiti marketed as "mid tier" ?
Basing GK104 as midrange based on "GK110" (which may be mythological for as much as anyone actually knows about it) being enthusiast/top-tier ? Making an assumption from an assumption ?
Whose to say that "GK110" isn't GTX 780...assuming it's anything at all (remember the 96TMU/384SP GT212). You know for a fact that "GK110" is of the same family/release as "GK104"?
No matter what, they don't have anything faster for now. So if what they do have is faster than the competition they name it accordingly and I guess that they price it accordingly, although we dont know the price yet, so making assumptions is stupid. Like Crap Daddy said, we are probably screwed, but that's something I've been saying the minute I saw HD7970's performance, but of course I was flamed for that.
So there is still a chance because NV has to fill some parts all the way down to the GTX560Ti/HD7850 which are 200-250$ cards and they have only one chip ready. The GK106 is nowhere to be seen (in fact it's more mysterious than the GK110).