Nvidia GT300 Yields under 3%

W1zzard said:

any time you read the words "yield" and "%" together in an article on the internet, click the X, no matter the source. Yes, not even press releases from AMD, Intel, NVIDIA, TSMC or God himself. Besides, that number has no meaning for you.

If you buy some coke do you ask how many plants died because the DEA burnt them down? No, you ask how much it costs and what it does for you (and how you can get more of it)

Click to expand...

Not only dose W1zzard get to benchmark the latest ATI cards..

He also gets to benchtest the latest coke stright from columbia

[Ion] said:

I won't believe it until I read it elsewhere, Charlie's articles are usually complete BS

Click to expand...

InTeL-iNsIdE said:

the same Charlie Demerjian who used to blog at the Inq and now blogs over at www.semiaccurate.com ? thos 2 websites say it all

Click to expand...

El Fiendo said:

Again, Charlie Demerjian. You can trust anything he says about nVidia about as far as I can fly with my arms. Also, note that the only places reporting it are simply quoting him. No one else mentions anything about this that isn't using his words.(...)

Click to expand...

Charlie may be a sensationalist anti-journalist, but he was spot-on with the GT212 production issues, which makes me believe he actually has a source in TSMC itself.

erocker said:

He is biased, but not necessarily incorrect. I still haven't seen anything on any Nvidia 40nm parts. To put it more truthfully it should be said that TMSC can't make Nvidia 40nm GPU's.

Click to expand...

This isn't true.
The Acer Aspire 5739G has a GT240M (GT216) and the Medion P7612 has a G210M (GT218). Both are already in the market for some time and both have a 40nm DX10.1 nVidia GPU.

Nonetheless, given that there's only one laptop with the 48sp GT216 and another one with the 16sp GT218, I'd say the yields for nVidia's GT21x are still not good to this date.
The 96 shader, GDDR5 GTS250/60 (GT215) is yet to be seen at all.

W1zzard said:

any time you read the words "yield" and "%" together in an article on the internet, click the X, no matter the source. Yes, not even press releases from AMD, Intel, NVIDIA, TSMC or God himself. Besides, that number has no meaning for you.

If you buy some coke do you ask how many plants died because the DEA burnt them down? No, you ask how much it costs and what it does for you (and how you can get more of it)

Click to expand...

Factory yields determine the profit margins on the waffer and on each sold GPU, which in turn has an important role in the final price of the graphics card.
Very bad yields could even lead to the cancellation of an unnanounced GPU, like what happened with the GT212, or even the cancellation of an anounced graphics card, as we saw with the Mobility HD4860 in the Asus K series.

hat said:

What about techpowerup?

Click to expand...

yes even tpu, we would never post such guesses anyway because any source can not be trusted

HalfAHertz said:

That's really hard to believe. He states, that TSMC produces 104 dies out of 1 waffer. If I'm not mistaken, they used 300mm waffers. Dunno if my mad math skillz are correct but that would leave us with a total area of ~7050mm^2 per waffer and a die area of over 650mm^2. That's supposedly twice the die area of the HD5870 and highly unlikely case...

Click to expand...

that is about rite, thats how big these are.

without a more reliable source, I find this hard to believe

I have not payed too much attention to the posts or the editorial (he doesn't even have a clue about what a news article is) itself, but come on, Demerjian really? We have to believe anything?

The very first day he links to a source that is not an editorial made by himself previously, I will start considering that anything he said might have a 3% of truth into it. I do read him a lot for fun, but when it comes to Nvidia he is so pathetic. It's so usual to read "as we found out some weeks ago" in blue letters ---> link to an editorial made by him in which the only link you can find is again to one of his previous editorials. You can travel as much as 20 links back this way, until you find one editorial which will not have a link to another of his editorials. Yes, you guessed it, there's no source or link at all. You finally found the one first editorial that came straight from his ass.

I still remember that thorought article in which he calculated how much every GTX2xx card costed Nvidia to make. Component after component, the final cost turned out to be in excess of $300 for the GTX260 and a lil more for the GTX280, before the distributors or retailers had even eaten their piece of the cake. According to him, then (that was when RV770 launched or around that date), Nvidia couldn't sell the cards below $400 and $550 respectively. Yeah, 1-2 weeks later stores were selling the GTX260 below manufacturig costs (calculated by him, thats' $300) and 2 months later they were losing more than $100 per card sold instead of earning the usual $40-50 per card, and they are selling for $140 now (wow to the cost reductions!!), yet Nvidia has always turned a profit and still had $200 millions to pay the bumpage thing. Magicians?

I do have to give him some credit though , because in the art of lying he is a master. You know, that art is based in that the perfect lie is the one that has some small percentage of truth. This way Nvidia might have some problems with GT300, so you just say 3% yields and just sit back watching the site hits go up.

There's no difference between him and the sensationalist media, where they don't mind to say that <insert famous actress> is pregnant, based on info coming from their ass in order to bump their TV share and when it comes to be false, no worries because "we will get an even bigger share when we say that we were the only ones that knew it was not truth, even if we were the ones that said she was in the first place." Who cares about truth when you can just make money. Literature sells and it neither is true isn't it? The blame is on those who believe anything he says, he didn't ever say it was true anyway...

W1zzard said:

any time you read the words "yield" and "%" together in an article on the internet, click the X, no matter the source. Yes, not even press releases from AMD, Intel, NVIDIA, TSMC or God himself. Besides, that number has no meaning for you.

If you buy some coke do you ask how many plants died because the DEA burnt them down? No, you ask how much it costs and what it does for you (and how you can get more of it)

Click to expand...

But my tpu tabs would go too

tigger said:

More interesting info-
http://www.semiaccurate.com/2009/08/13/gt300-have-nvio-chip/

Click to expand...

Considering that site is decked out with three ATI/AMD adverts and says silly statements like "Nvidia, don't let the door hit your ass on the way out". I would say you have another cooked up article made by a basement dwelling idiot.

*EDIT* I now see Benetanegia has weighed in on the subject. I agree with you for the most part. I hate it when blogs without sources are respected without proof.

If this would be true, it could even tha market balance, which would be good.

mah i wouldn't beleive any of it, the guy who wrote it loves amd, and probably posts himself an amd christmas card,
pretending it's from amd saying they love him
anyway to nvidia

They're merely making engg. samples, so obviously they're going to pick the best ones on the wafer. GT300 will be an engineering marvel on 40 nm. They're not facing foundry/yield issues during mass-production, so Charlie's point doesn't hold any ground.

After reading this I wonder what kind of bad buzz this generates for anything except stock holders. Really it doesn't mean that nV's chips suck, or definitively that their product would hypothetically take more time to be released. What it does mean is that an unreliable source is giving figures that he may or may not have access. Fabricating samples... it's soooo early in the game I doubt this sort of message true or false means much to us, the end users.

Benetanegia said:

there's no source or link at all

Click to expand...

I've read the 2nd paragraph... not to spam the ATI HD 5800 Series Leaks article so I'll post here, this line isn't really accurate.

What exactly do you expect from a source statement?... if he has insiders, or a few contacts that don't want to be mentioned (because they will probably get in trouble?) that won't be suffice to discredit him!

btarunr posted some articles in this manor as well... his sources... his privileges!

The thing is that sometimes you can relay the already printed / posted news while other times you have to rely on your contacts and be "the first"... among other firsts.

I don't say he's a saint and that's he's right... nor don't I agree of his hatred towards nvidia, but neither can I on yours towards him or AMD's ads... which BTW point out to expired newegg offers.

What Easo says is what I want so see... thus I would like it to be the way Charlie said.

And my point about GT300 being to expensive and having a low yield to actually be considered as a real product still stands. Good for them that the GT300 will be faster than anything ATI has... in my view, it's more of a PR stunt which reminds me of Phenom's debute.

If I would make the usual comparison... with cars... Ferrari, Lamborghini, Maserati do make some really fast limited-edition cars ... but they were also threatened by bankruptcy until someone who produces mainstream cars bought them out... else they would have faded. We don't want that for Nvidia, nor do we want them to always be the best!... the ACTUAL BEST!

Andy77 said:

I've read the 2nd paragraph... not to spam the ATI HD 5800 Series Leaks article so I'll post here, this line isn't really accurate.

What exactly do you expect from a source statement?... if he has insiders, or a few contacts that don't want to be mentioned (because they will probably get in trouble?) that won't be suffice to discredit him!

btarunr posted some articles in this manor as well... his sources... his privileges!

The thing is that sometimes you can relay the already printed / posted news while other times you have to rely on your contacts and be "the first"... among other firsts.

I don't say he's a saint and that's he's right... nor don't I agree of his hatred towards nvidia, but neither can I on yours towards him or AMD's ads... which BTW point out to expired newegg offers.

What Easo says is what I want so see... thus I would like it to be the way Charlie said.

And my point about GT300 being to expensive and having a low yield to actually be considered as a real product still stands. Good for them that the GT300 will be faster than anything ATI has... in my view, it's more of a PR stunt which reminds me of Phenom's debute.

If I would make the usual comparison... with cars... Ferrari, Lamborghini, Maserati do make some really fast limited-edition cars ... but they were also threatened by bankruptcy until someone who produces mainstream cars bought them out... else they would have faded. We don't want that for Nvidia, nor do we want them to always be the best!... the ACTUAL BEST!

Click to expand...

There's a substantial difference between protecting your sources and what Charlie does. If you take Fudzilla for example (another rumor spreader), they never tell their sources, but they limit themselves to giving news. They don't write editorials, giving their opinion on something he can't give proofs about and when they are found wrong, they retract.

It's as simple as that, you have no proofs for whatever reason, you shut your mouth up. But that's not the problem, anyway, the problem is that 99% of what those sources tell him about Nvidia ends up being false. There are plenty of examples of that and just one in which he was correct. There's a difference though, when he is correct, he is not the only one saying it and you have Nvidia making some sort comment on it the day after it's made public. That's what happens when you deal with true info vs false info. I don't know if he makes things up, or if he is naive enough to constantly believe a "source" that makes things up, either way, 99% of what he says comes from inside one ass.

Benetanegia said:

There's a substantial difference between protecting your sources and what Charlie does. If you take Fudzilla for example (another rumor spreader), they never tell their sources, but they limit themselves to giving news. They don't write editorials, giving their opinion on something he can't give proofs about and when they are found wrong, they retract.

It's as simple as that, you have no proofs for whatever reason, you shut your mouth up. But that's not the problem, anyway, the problem is that 99% of what those sources tell him about Nvidia ends up being false. There are plenty of examples of that and just one in which he was correct. There's a difference though, when he is correct, he is not the only one saying it and you have Nvidia making some sort comment on it the day after it's made public. That's what happens when you deal with true info vs false info. I don't know if he makes things up, or if he is naive enough to constantly believe a "source" that makes things up, either way, 99% of what he says comes from inside one ass.

Click to expand...

Yeah.. speaking of fud,

Several people confirm
http://www.fudzilla.com/content/view/15535/34/

Nvidia is still not revealing any details about its chip that we still call GT300 but one thing is getting quite certain. A few people who saw just leaked performance data on ATI's soon to launch Radeon HD 5870 have told Fudzilla that they are absolutely confident that GT300 will win over Radeon HD 5870.

Since GT300 has its GX2 brother they should not fear much against Radeon 5870 X2 whenever that comes out, as if Nvidia wins a single chip battle, they can win the dual one as well.

We can only confirm that GT300 is not a GT200 in 40nm with DirectX 11 support. It’s a brand new chip that was designed almost entirely from the ground up. Industry sources believe that this is the biggest change since G80 was launched and that you can expect such level of innovation and change.

The launch as far as we know is still slated for late November and hasn’t been confirmed.

Click to expand...

Bjorn_Of_Iceland said:

Yeah.. speaking of fud,

Several people confirm
http://www.fudzilla.com/content/view/15535/34/

Click to expand...

I don't see several people confirm anywhere there.

And again there's a signficant difference between saying some people have told fudzilla bla bla bla, and leave it there (they inform, they pst news), to starting by saying that, and then taking that as granted and if it was a fact, continue saying that because of bla, XXXX comapny is going to bla, and is a joke and bla, and patethic bla.

Fudzilla informs about what their sources think. States his sources are confident about one thing, they don't make that theirs and utilise it later as they better think will help them gain some "clicking share".

For example when any site gets the info "GT300 will be 500 mm2", including Fudzilla, they post: We've been told that GT300 will e 500mm2" and maybe say a little bit more. When Charlie gets that, he goes on to produce a melodramatic editorial about how the chip will cost twice to produce (which until now is mostly true, althought there are far more things to consider that he simply forgoes), how that will make the card twice as expensive (whaaat? how?) and only will be able to compete with the X2 in price, and hold on to your seat because he didn't finish yet: GT300 will NOT be faster than a single RV870, because... ah no there is no a reason for that, he doesn't say why exactly. He does link you to another editorial made by him in which based on the fact that Nvidia is making a new architecture using MIMD, it will fail. IN that linked editorial, the news are "Nvidia to use MIMD", and out of that he concludes it will be slower. Well from now on GT300 is slwer and he will happily say that many times in the future with his precious link at hand in case someone doubts the "well stablished fact" (sarcasm) that GT300 will be slow and primarily that Nvidia simply sucks in everything, because... well he has enough links to probe that several times...

I believe it will just follow current market trends until you see a good established reason to believe otherwise.

I would predict the GT300 to be faster and more expensive. The ATi offerrings will all be most likely cheaper at their respective price points. It doesn't make one company better than the other. I enjoy offerings from both. I have preferred Nvidia more lately, but competition needs to stay strong.

I will be waiting patiently for the GT300, only way I go ATI is if a new Folding Client is released