NVIDIA Kepler speculation thread

[DavidGraham]

Why are we even discussing this non-sense? , we are getting derailed off the important technical bits here for useless -so called- issues that have no impact whatsoever on the product the user receives .

For all of their alleged continuous problems , NVIDIA seems to be doing an admirable job of staying profitable and sustaining a strong business portfolio and excellent innovative competitiveness , On the other hand , AMD allegedly and miraculously (always) trouble free , is stumbling and losing money on multiple fronts while having less market penetration where it counts .

So unless you guys have answers for these dilemmas , I suggest we get back to technical discussion , and leave the mumbo jumbo stuff to Mr.Charlie and his goons , together they could use another healthy dose of useless rumors to brag about .

 

[trinibwoy]

No, you can do a lot more than that. If that was NVIDIA's attitude then I can see why they would be complaining now.

Really? Is nVidia is the only design firm that has encountered problems at the foundry? Maybe they should borrow AMDs crystal ball

How, exactly, do you think the supplier can underperform?

Is that a serious question?

Step 1: Make promises
Step 2: Don't meet them

 

[AlphaWolf]

Really? Is nVidia is the only design firm that has encountered problems at the foundry? Maybe they should borrow AMDs crystal ball

Or perhaps their engineers.

 

[no-X]

As for TSMC, their 28nm wafer production in Q1/2012 exceeded plans. And as for nVidia...

Our transition to 28nm is going better than 40nm, and yields are better than our original plan.

Yields were better than expected, manufacturing capacity was better than expected and despite that, they are complaining.

 

[hatter]

Qualcomm has publicly complained about 28nm capacity at TSMC. It says that it can't meet the demand for new Snapdragons. May be Nvidia has some other issues, but it is wrong to say that it is the only company having problems.

http://www.bit-tech.net/news/hardware/2012/04/19/qualcomm-28nm-capacity/1

And for whatever it is worth, I feel that despite bad pricing the availability HD 79XX cards has improved only now, five months after they were introduced.

 

[Dave Baumann]

Yields were better than expected, manufacturing capacity was better than expected and despite that, they are complaining.

That had changed by CC in Feb:

http://seekingalpha.com/article/370...2-results-earnings-call-transcript?part=qanda

Jen-Hsun Huang

The top line decline for Q1 is expected to be due to the hard disk drive shortage continuing, as well as a shortage of 28-nanometer wafers. We're ramping our Kepler generation very hard, and we could use more wafers. The gross margin decline is contributed almost entirely to the yields of 28-nanometer being lower than expected. And that is, I guess, unsurprising at this point. And because we have -- we use wafer-based pricing now, when the yield is lower, our cost is higher.

Be interesting to see the comments on Friday.

 

[3dcgi]

I think you completely missed the point here.

The problem is with TSMC, who is not able to meet the demand for 28nm wafers.

Graphics card shortages to see improvements in late May
May 3, 01:20
AMD and Nvidia, impacted by Taiwan Semiconductor Manufacturing Company's (TSMC's) fully-booked capacity, had rather weak shipment performance in the first quarter; however, as more capacity will be gradually released by TSMC, shortages of 28nm graphics cards are expected to improve in late May, according...



Actually it was TSMC who had all sorts of problems with the 40nm manufacturing process, resulting in very poor yields the entire second half of 2009. This is why ATi HD5xxx gpu's were extremely hard to find and nVidia's Fermi had to be delayed because of its complexity and size... and horrible yields.

I'm not missing any point. If TSMC's yields are ramping slower than expected it will certainly affect availability, but you can't attribute hard to find HD5xxx's to that alone. Most yield issues were ironed out by then.

When the economy was bad TSMC stalled their build out of 40nm, some articles said they temporarily laid off workers. Part of the reason for stalling the build out of 40nm capacity was customers like AMD didn't project accurate demand. By the time the demand was obvious it took a while for TSMC to meet demand.

 

[Alexko]

Qualcomm has publicly complained about 28nm capacity at TSMC. It says that it can't meet the demand for new Snapdragons. May be Nvidia has some other issues, but it is wrong to say that it is the only company having problems.

http://www.bit-tech.net/news/hardware/2012/04/19/qualcomm-28nm-capacity/1

And for whatever it is worth, I feel that despite bad pricing the availability HD 79XX cards has improved only now, five months after they were introduced.

But:

Qualcomm's chief executive Paul Jacobs told press and analysts during the company's earnings call late yesterday. 'Although the manufacturing yields are progressing per expectations, there is a shortage of 28nm capacity.

 

[A1xLLcqAgt0qc2RyMz0y]

That had changed by CC in Feb:

http://seekingalpha.com/article/370...2-results-earnings-call-transcript?part=qanda

Be interesting to see the comments on Friday.

Yes it will be interesting to hear if there is any improvement in yields.

---------

More quote(s):

And so we've transitioned to a wafer-based pricing for some time. And our expectation, of course, is that the yields will improve as they have in the previous generation nodes, and as the yields improve, our output would increase and our costs will decline. And that's why we expect to exit the year at 52% or about.

----

TSMC is doing fabulously with 28, and this year, with 28 relative to 40, it's surely a huge improvement. But during transition years, because we have so much of our business tied to the leading edge nodes and because when we transition GPUs, we transition so much of it at once. And this year -- particularly unique this year is that the number of notebook design wins that we have and the number of 28-nanometer GPU design wins that we have is, at OEMs, is much, much higher than our previous generation. And so all of the transition of 28 is going to be very fast. And so we're just going to have to continue to work with TSMC and get the yields of 28-nanometer up as fast as possible. And we surely expect that by the end of the year, we're going to be in a pretty good place. We're in a pretty decent place now, but we just need to get the yields up.

 

[rpg.314]

AMD not complaining is an indicator of pretty much nothing....

It means their yields are fine. It means they forecast their demand correctly and TSMC was able to meet their demands. BTW, Qualcomm's yields are fine too.

 

[trinibwoy]

It means their yields are fine. It means they forecast their demand correctly and TSMC was able to meet their demands. BTW, Qualcomm's yields are fine too.

It means nothing. You are drawing all sorts of far reaching conclusions based on the absence of noise. I have a couple hundred clients. The amount of noise one of them makes typically has nothing to do with the severity of the problem.

Show me high demand for AMDs products and correspondingly high output from TSMC. Anything else is spurious nonsense.

 

[iMacmatician]

Bright Side of News: "AMD Counter-Attacks GeForce GTX 670, Questions NVIDIA's Kepler Availability."

NVIDIA may be launching a GTX 670 in a couple of hours, but that did not stop a PR representative from AMD to publicly question the availability of NVIDIA hardware. Instead of providing a comment, we have decided to leave the email "as is" and leave you to comment on the subject.

 

[rpg.314]

It means nothing. You are drawing all sorts of far reaching conclusions based on the absence of noise. I have a couple hundred clients. The amount of noise one of them makes typically has nothing to do with the severity of the problem.

TSMC has how many high volume 28 nm clients? AMD, NV, Qualcomm.....????? When the number of clients can be counted on finger tips, then the noise one makes or does not makes matters. A lot.

Show me high demand for AMDs products and correspondingly high output from TSMC. Anything else is spurious nonsense.

Based upon public information, only nv is having yield problems at TSMC. Everybody else's 28 nm yields are fine. Show me other 28 nm clients complaning about yields. Otherwise this is spurious non sense.

 

[UniversalTruth]

He's not missing the point. Demand is planned. If, all of a sudden, when the products are shipping someone is saying that demand is not being met then it is more likely that something has gone awry with the planning (i.e. it was either not planned accurately or there are yeild issues).

Once again, not everyone is complaining.

NV's situation is particularly interesting. GK104 is a midrange chip and as being such the planned demand should have been somehow higher with respect to current demand. Obviously now the chip covers entirely different segment which is characterised with (much) lower demand. So despite this and the lower market requirements they suddenly are still unable to supply enough of them. Disastrous yield maybe.


If I remember correctly, there was a moment (maybe late last year) when NV cancelled some wafer orders.

 

[madyasiwi]

Based upon public information, only nv is having yield problems at TSMC. Everybody else's 28 nm yields are fine. Show me other 28 nm clients complaning about yields. Otherwise this is spurious non sense.

Overshooting their expectation and miscalculated their production capacity plan is very different than "only nv is having yield problems at TSMC".

And don't forget that no other client bear the pain of making 600 mm2 parts.

 

[DavidGraham]

The way I see it , AMD feels threatened by NVIDIA's latest offers , so they play the availability card , They poke Charlie to write some of his "glamorous" pieces , and they have Mr.Dave here in the forums raising the same questions and doing pretty much the same thing , albeit with much more elegance .

That is really the only thing that explains to me why we are even discussing this absolutely useless and trivial issue .

 

[xEx]

Apparently this is the first review of the 670, so enjoy ;D.

http://www.ozeros.com/2012/05/review-nvidia-geforce-gtx-670/

Its in Spanish but number are numbers and talking about they, its funny the way the 580 is almost at the 7970 performance(specially on RD5 where in Kitguru´s 2 months old review the 580 is below the 7850 and I don't think the OC it has could scale that much 0.0 ) and the 670 is head to head to the 680.

 

[Bouncing Zabaglione Bros.]

Some posters in this thread don't seem to understand the basic difference between "yield" and "capacity".

AMD and Qualcomm appear to be experiencing good yield but low capacity. Nvidia are saying they are seeing poor yield and low capacity. The question is why the difference.

 

[jaredpace]

TR's review of the 690 is up. Nvidia made some progress with the frametime metering fix for microstutter. It's looking better for the 690 compared to all other AMD/NV multi-gpu setups, however, still not as consistent as single-gpu. Still a marked improvement. GJ Nv!

 

[trinibwoy]

TSMC has how many high volume 28 nm clients? AMD, NV, Qualcomm.....????? When the number of clients can be counted on finger tips, then the noise one makes or does not makes matters. A lot.

Arbitrary correlation that says nothing. AMD could be quiet for any number of reasons - including that they overestimated demand. How do you know the real reason without numbers to back it up?

Based upon public information..

What public information?