NVIDIA Kepler speculation thread

Nvidia won't waste the chance to milk the 660 name for all it's worth.

If there are two different chips, the faster 660 Ti will be GK104 and the slower 660 will be GK106, with 660 SE for the salvage part of the GK106. This let's them ride on the success of the 660 Ti being faster than the 7870, while the actual 660 (GK106) is slower but almost certainly more expensive. I'd put money on this.

They did this with the 560, they'll do it here. If Nvidia excels at anything it's confusing their buyers and they'll lap up the slower 660 based on the prowess of the 660 Ti.
Huh? This is illogical. Besides, the 560 offered better performance per dollar than the 560 Ti when it launched.
 
Huh? This is illogical. Besides, the 560 offered better performance per dollar than the 560 Ti when it launched.

Slower than the 7870 I mean.

This is the same situation as the 560 Ti and 560 - the 560 was slower than the 6870 (at least on release), but more expensive. Nvidia get away with it because of the 560 Ti being a bit faster. It's creating a halo effect.

They'll do the same thing here - the 660 sans-Ti will be slower than the 7870, cost more, and probably sell more units. It'll also be reviewed with overclocked cards just to reemphasise the superiority that doesn't actually exist. Nvidia knows every trick in the book and has been using them all for the past 2 years.
 
Slower than the 7870 I mean.

This is the same situation as the 560 Ti and 560 - the 560 was slower than the 6870 (at least on release), but more expensive. Nvidia get away with it because of the 560 Ti being a bit faster. It's creating a halo effect.

They'll do the same thing here - the 660 sans-Ti will be slower than the 7870, cost more, and probably sell more units. It'll also be reviewed with overclocked cards just to reemphasise the superiority that doesn't actually exist. Nvidia knows every trick in the book and has been using them all for the past 2 years.
I'd argue that Nvidia gets away with it because they're Nvidia. They have better brand recognition and a larger number of drooling fanboys. It'll catch up with them eventually, but they're enjoying the ride while they can.
 
Nvidia has the better brand recognition

They have better brand recognition and a larger number of drooling fanboys.

Yes, Nvidia has the better brand recognition. Most of Nvidia's sales are to OEMs not to drooling fanboys.

It'll catch up with them eventually, but they're enjoying the ride while they can.
I have heard this quote over and over again from the RED Team fanbois and year after year it doesn't happen. And it won't.
 
I'd argue that Nvidia gets away with it because they're Nvidia. They have better brand recognition and a larger number of drooling fanboys. It'll catch up with them eventually, but they're enjoying the ride while they can.

Yes I agree, however having some tricks up your sleeve when you're slower also helps. I still find it utterly bizarre that so many lost their head over the 560 Ti, which was on par with the 5870 (with the 560 Ti more expensive of course) some 16 months later. Only genius marketing can get away with that.
 
There's a finite number of wafers available. Ordering "enough wafers" can't change that if that number is higher than TSMC's available capacity.

Of course, but at some point of planning for new 28nm line TSMC asked their customers how much capacity they will need initially.
You don't build very expensive factory if no one will utilize it's capacity and you also try to build enough capacity to fulfill all orders.

It might be TSMC couldn't afford to build enough capacity, but with order guarantees from partners they would get loan required. Maybe 28nm tool makers can't supply more equipment but I doubt that.

Without inside knowledge we can put arguments and counter-arguments to almost any possible situation. What I'm trying to say here, reality is usually more gray than black&white so the blame is probably somewhere in the middle between TSMC, their clients and tool manufacturers.
 
Without inside knowledge we can put arguments and counter-arguments to almost any possible situation. What I'm trying to say here, reality is usually more gray than black&white so the blame is probably somewhere in the middle between TSMC, their clients and tool manufacturers.

Yep, 100% agreed.
 
Of course, but at some point of planning for new 28nm line TSMC asked their customers how much capacity they will need initially.
You don't build very expensive factory if no one will utilize it's capacity and you also try to build enough capacity to fulfill all orders.
It doesn't work that way. Building a fab and getting is fully up and running is a multi-year endeavor. There's no way TSMC can ask its partners to predict years up front exactly how many wafers need to be allocated.

Any fab that TSMC has built before has been used at full capacity, and given the fact that they skipped 32nm and went straight to 28nm, there was no question that there'd be massive pent-up demand for it. No need to first ask around. It just takes time to fine-tune a new process and you don't go all out with a new process before you've gotten some initial indications that your recipe is going to work.

It might be TSMC couldn't afford to build enough capacity, but with order guarantees from partners they would get loan required. Maybe 28nm tool makers can't supply more equipment but I doubt that.
TSMC is not the kind of company where a few hundred million dollars more or less of customer commitments are going to make or break getting a loan from their local credit union. ;)

Without inside knowledge we can put arguments and counter-arguments to almost any possible situation. What I'm trying to say here, reality is usually more gray than black&white so the blame is probably somewhere in the middle between TSMC, their clients and tool manufacturers.
It's really quite simple: there are now more players who need state-of-the-art silicon during the early days of the process, so there's a shortage, which make Qualcomm look elsewhere. They'd be stupid not to.
 
I don't think this has been mentioned in this thread before, but there are new (as in announced last month or so) mobile Quadros: K500M, K1000M, K2000M, K3000M, K4000M, and K5000M. I would assume that the K4000M (960 CCs, 256-bit bus) is a disabled GK104, but the interesting thing is, K3000M is 576 CCs and 256-bit bus. Would that be a GK106 (with presumably 1 SMX disabled) or a GK104 with 5 SMXs disabled?
 
I don't think this has been mentioned in this thread before, but there are new (as in announced last month or so) mobile Quadros: K500M, K1000M, K2000M, K3000M, K4000M, and K5000M. I would assume that the K4000M (960 CCs, 256-bit bus) is a disabled GK104, but the interesting thing is, K3000M is 576 CCs and 256-bit bus. Would that be a GK106 (with presumably 1 SMX disabled) or a GK104 with 5 SMXs disabled?

All above the K2000M are N14E chips, so I assume cut down GK104s.
 
well, if 1152CCs and 192 bit is truth for full blown GK106 not 768CCs/192bit then something should have happened like Barts.. there were 2 Barts and one was canceled.. NV said they were disappointed by Tahiti but i think they wouldnt expect Pitcairn will be that strong too.. it would give hints about long delay as well..
 
Sorry for the self quote, but Steam seems to confim those numbers for their june data set.
Well, not being steam surveys and thus having a rather limited number of participants are two quickpolls in german enthusiast forae are those two quickpolls. But apart from having a strong tendency of over emphasizing the high-end portion of the market, both communities see GTX 680 roughly on par with HD 7970 despite being on the market for a significantly shorter amount of time.

http://www.forum-3dcenter.org/vbulletin/showthread.php?t=528393
http://extreme.pcgameshardware.de/q...nutzt-du-derzeit-zum-spielen-juni-2012-a.html
NVIDIA GeForce GTX 680 - - 0.31% 0.56% 0.74% +0.18%
ATI Radeon HD 7970 0.31% 0.38% 0.41% 0.50% 0.50% +0.00%

This does definitely not look like no one out there wasn't able to get a 680 in the past month.
 
Man from Atlantis: I don't quite believe the story about much higher expectations regarding Tahiti's performance. GF104 "done right" (fully enabled and with GF114-like clocks) would be also as fast as HD 5870. Both GF104 and GK104 were targeted relatively to the same performance level as competitiors high-end GPU. It's hard to believe, that Nvidia would be surprised just from one of these GPU.
 
Man from Atlantis: I don't quite believe the story about much higher expectations regarding Tahiti's performance. GF104 "done right" (fully enabled and with GF114-like clocks) would be also as fast as HD 5870. Both GF104 and GK104 were targeted relatively to the same performance level as competitiors high-end GPU. It's hard to believe, that Nvidia would be surprised just from one of these GPU.

:LOL:

WoW, sorry to disturb, but I had much higher expectations about Tahiti and to me it is underwhelming. Why wouldn't NV think so too, without of course these technical details you are trying to protect your point? They are meaningless from time frame point of view.

:?:
 
Sorry for the self quote, but Steam seems to confim those numbers for their june data set.

NVIDIA GeForce GTX 680 - - 0.31% 0.56% 0.74% +0.18%
ATI Radeon HD 7970 0.31% 0.38% 0.41% 0.50% 0.50% +0.00%

This does definitely not look like no one out there wasn't able to get a 680 in the past month.

It is a little strange the 7970 gained no share, especially with price going quite down on standard editions (in Europe, you could find a SE 7970 for around 50-60€ less than the 680) and the GHz edition launch. As usual, these results should be taken with a grain of salt... (and, as usual, there is no sign of 670 and lower Kepler parts in the list...)
 
:LOL:

WoW, sorry to disturb, but I had much higher expectations about Tahiti and to me it is underwhelming. Why wouldn't NV think so too, without of course these technical details you are trying to protect your point? They are meaningless from time frame point of view.

:?:

There is simply no way in hell that Nvidia was "surprised" by Tahiti.
We knew almost exactly what Tahiti was about a year ago, July 2011.
Does anyone really believe Nvidia planned to compete against a full GPU family with a single ASIC for six months with a capacity constrained fab on a new process?

GK104 is Nvidia's RV670?
 
UniversalTruth: Well... We can't ignore, that GF104 had potential comparable to Cypress. Of course it came underclocked and partially disabled due to some complications, but that doesn't change anything on the fact, that GF104 was designed as a product offering some level of performance - in fact comparable to AMD's high-end.

The same applies to GK104 with the only exception - GK104 doesn't have GF104's issues. Performance target of GF104 and GK104 (relatively to AMD's high-end) is almost identical. It's really hard to accept, that Nvidia could be surprised in one case and not in the other one.
 
Back
Top