Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Nvidia announces the new Titan X, A Powerful $1,200 GPU
blueturtle13
Member LegendaryPosts: 12,698
Nvidia's breakneck GPU release schedule continues today with the
announcement of the
new Titan X, the company's most powerful graphics card yet.
The new Titan X, which Nvidia calls "the biggest GPU ever built," has 12 billion transistors in total — and you'll be paying about $100 per billion. The card will be available on August 2nd in the US and Europe for $1,200, with an Asia release forthcoming.
new Titan X, the company's most powerful graphics card yet.
The new Titan X, which Nvidia calls "the biggest GPU ever built," has 12 billion transistors in total — and you'll be paying about $100 per billion. The card will be available on August 2nd in the US and Europe for $1,200, with an Asia release forthcoming.
Comments
And frankly....25% faster than 1080....lot of people will be dissapointed since since they hyped 50%.
I called the price (well on 25% faster card than 1080 that was supposed to be 1080ti)....1199$
Also this is 99% GP100 chip
http://images.nvidia.com/content/tesla/pdf/nvidia-tesla-p100-PCIe-datasheet.pdf
It's a status symbol.
I wonder if they will have wide availability like they have with the 1080.
Also, only FE versions. I guess that is typical for the Titans though.
This serves three purposes:
1) It lets nVidia say they have the fastest card, and it's this fast. That's a big marketing point, and it's free advertising for the brand even if they sell 0 of these cards.
2) It's also for the investors. They see things like this, the nVidia stock price will move (which direction is anyone's guess, but nVidia would obviously like it to be up).
3) Those very very few select people who have the money to burn and want to chase benchmarks. This is analogous to the car people who buy Ferrarris and Lamborghinis. There aren't many of them, but us mere mortals all turn our heads when we see one on the road.
Please do not respond to me, even if I ask you a question, its rhetorical.
Please do not respond to me
traveller, interloper, anomaly, iteration
Please do not respond to me, even if I ask you a question, its rhetorical.
Please do not respond to me
Apparently Nvidia claims it has 12 billion transistors, while GP100 has 15.3 billion transistors. So they're not the same chip. Which means this one is really weird. Of course, for the new Titan X to have GDDR5X while GP100 has HBM2 would also be very strange if they were the same chip, as that's a ton of die space for completely independent memory controllers.
In the past, Nvidia and AMD have generally used the same chip to be their top end graphics and compute chip. GP100 has the fixed-function graphics hardware, so I figured they'd do that again, as that's purely wasted silicon in a compute chip. Apparently they won't. Sometimes that dual-purpose chip has gone heavier on compute (Nvidia Fermi GF100/GF110, AMD Hawaii) and sometimes been a mostly graphics part with a little bit of compute added (AMD Cypress and Cayman, Nvidia GK110 and GM200). But it looks like this time, they're making two huge chips: one pure graphics chip and one compute-heavy chip.
My guess is that Nvidia got GP100 back, decided it was broken enough to be unsuitable for a high end graphics chip, and so they then decided to make GP102, which "fixed" the problems by taking a bunch of compute stuff out. That's fine for graphics, but obviously not fine for Tesla cards, which have to go with GP100 whether it's broken or not.
This seems to be the season for paper launches, with still nothing in stock at MSRP, whether from Nvidia or AMD. Such a huge chip on such a new process node is sure to have problematic yields, though $1200 per chip can handle some yield problems and still be profitable. But I'd still bet on a paper launch.
With the GTX 1080, I said AMD could easily beat it just buy scaling up die size and power consumption. Nvidia did exactly that to create the new Titan X, and it's not at all clear whether AMD will be able to beat that by scaling up. And even if they can, it's not at all clear that they will, as AMD usually doesn't go for huge dies. Fiji was an exception, as it launched about 3 1/2 years after their first cards on the same process node, simply because there wasn't another process node ready to move to.
If you want the top end 14/16 nm card for gaming, this is it, or at least pretty close to it. It's probably a salvage part, and they'll likely later launch a fully functional part that is faster. But that's not going to be a huge jump in performance.
But $1200? Even if AMD cedes the $1200 consumer graphics card market to Nvidia entirely, is that really a large enough market to justify making a huge, low volume chip for it? Remember, they're not sharing it with Tesla cards this time, though there will presumably be Quadro cards on the same chip.
It's long been believed in some quarters that huge chips like this aren't primarily real products that they expect a lot of real customers to buy. Rather, they're halo products, with the intention that gamers read reviews and see that Nvidia has the fastest card, but $1200 is out of their price range. But they conclude that Nvidia must be better because Nvidia has the fastest top end card, and they are willing to spend $200, so they go buy whatever Nvidia card they can find for $200, and overpay for something stupid. I'm not sure how common that effect is, but Nvidia sure acts like they believe it's real, and ATI used to do so, too.
GTX 1080 is a roughly 300mm2 chip, this is probably roughly 600mm2 with the exact same overall architecture as GP104
"The surest way to corrupt a youth is to instruct him to hold in higher esteem those who think alike than those who think differently."
- Friedrich Nietzsche
No matter how wealthy I was, dropping $1200 on a card that will be 'third fastest' in less than two years makes absolutely no sense.
The world is going to the dogs, which is just how I planned it!
http://www.kidsmmorpg.com/
The Tesla C870 used the same chip as the GeForce 8800 GTX.
The Tesla C1060 used the same chip as the GeForce GTX 280.
The Tesla M2070 used the same chip as the GeForce GTX 480.
The Tesla M2090 used the same chip as the GeForce GTX 580.
The Tesla K10 used the same chip as the GeForce GTX 680.
The Tesla K40 used the same chip as the GeForce GTX 780 Ti.
The Tesla M6 used the same chip as the GeForce GTX 980.
The Tesla M40 used the same chip as the GeForce GTX Titan X.
The FireStream 9170 used the same chip as the Radeon HD 3870.
The FireStream 9270 used the same chip as the Radeon HD 4870.
The FireStream 9370 used the same chip as the Radeon HD 5870.
The FirePro S9050 used the same chip as the Radeon HD 7970.
The FirePro S9150 used the same chip as the Radeon R9 290X.
The FirePro S9300 X2 used the same chip as the Radeon R9 Fury X.
There is one oddball exception, for the Tesla K80, which used a slightly different chip from the Tesla K40, which reduced the number of compute units, reduced the clock speed, doubled the register file size per compute unit, stuck two of them on a board--and didn't launch until Nvidia had otherwise moved on to Maxwell.
But here, it looks like Nvidia is making a high end graphics chip that won't be used for compute, and a separate high end compute chip that won't be used for graphics, and at about the same time. Unlike with the Tesla K80, in this case, the huge compute chip and the slightly less huge graphics chip are of very different architectures.
Now, there's nothing unethical or illegal about doing this. It's kind of like how Intel could create a completely custom die for the Xeon E3 rather than just reusing one from laptop and desktop quad cores. Or they could create a completely new chip for their desktop-E series rather than just using one that is also there for Xeon E5. But they don't, as it's too expensive to design and manufacture for too little to gain.
Nvidia apparently decided that they would eat that cost and make separate chips. Again, there's nothing illegal or unethical about this, but in the past, they've always shied away from it due to the cost.
[mod edit]
Somebody, somewhere has better skills as you have, more experience as you have, is smarter than you, has more friends as you do and can stay online longer. Just pray he's not out to get you.
This doesn't follow the pattern (or as some people like to call them, statistics). It seems a very high cost effort to introduce a card that may sell in the single-digit thousands. Of course, no one is sharing actual sales numbers, but given that no Titan model at all even pops up on the Steam Hardware survey, it isn't that big.
But, the x80Ti models do show up - the 980Ti is at around 1%, and the 780Ti still registers at 0.25%
My guess - HBM2 is not looking promising, and/or GP100 is just very borked (which is what Quiz suspects), and/or (and this one is a long shot) yields/production/whatever of GP104 are just so bad they needed to introduce these now to take the pressure off that inventory. - those are my guesses. It takes longer than a few months to spin up a new die, so for the GP102 to be out on a card (even a card that will be, by all indications, extremely low volume), this had to start months ago.
Any which way, it means that a GP100 based Titan/1080Ti wasn't going to work out in time to make this generation, or (and I doubt this) this was nVidia's master plan all along. Now, Titan's don't move a lot of cards, but the high end TI does move some, and if they share the same die, then you don't get one without the other. And if you aren't quite sure how your yields are going to be yet - release the most constraining part first at a prohibitive cost, and see what happens...