GF100 (Fermi) previews and discussion

Page 18 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
They not only didnt make a profit, their inventories are climbimg, their stock values are shrinking, they dont have anything competitive with pricing or quantities in the mig higher ends, the low ends are all old old tech, except the overpriced stuff, theyre dumping using the 330, so, no telling what youll get with that model, beats key chains I guess, and until they make something competitive, theyll continue to slide downhill
 
Nah, nVidia still has a solid low-range line up to get some mone. So they might not get the big bucks, but they're not going out of business either. Besides, they could just do the "IT'S REBRANDING TIME".

The GTS250, GTS220 and GTS210 are very good contenders at their price ranges if you ask me. And notebook wise, at least not thinking about performance ones, they still have a very solid ground I'd say.

Like I said on other topic, Fermi 1 will be the HD2900 of nVidia in the long term fight.

Oh, and that Intel is looking at this very closely 😛

Cheers!

EDIT: Typo.
 
Their excuses are the lack of 40nm for those lower ends, and thats never been a problem for ATI, so once again, its nVidia saying it this orthat, they have no design wins, and are losing them quickly to ATI, especially in mobile, where theyll find themselves way behind very shortly.
Im think Intel will continue to watch, until next gens out anyways
 
But notebook wise, if you look closely JD, ATI has the upper hand on performance notebooks, wich are like the 10% (at best?) of the whole picture (I have no solid info on that one, so feel free to correct me 😛), and mainstream wise, DX11 is of no use (5570 anyone?) at all. The only thing on the lower end wich comes into play is HD playback capabilities in first place and prolly other "add ons" you can find nice for each card (UVD2, for example).

But with that assumption, even IF Fermi turns out to be a major fail-chip (wich I'd say it prolly won't) they can revise it, change it, and come clean again supporting R&D with the lower end market (notebooks included) revenue. Remember the sales figures are always balanced towards the mid-low end (or tight budget) crowd.

I have to cheer for nVidia on this one, so I'll go get my green jacket out of the basement, rofl.

Cheers! xD
 


I'm pretty sure they did make a profit, and all I've heard about inventories is that they don't have enough anymore..... nVidia was also predicted to regain some of the market share it lost to Intel in the next couple of months (I can find the article for that if you want, but ain't putting in the effort unless I have to), where as ATI lost market share even with the 58xx series cards due to lack of supply anyways.

nVidia isn't in a bad place at all... yet. If Fermi fails, or they delay it another few months, then they are in a bad position. They are pretty much breaking even right now and Tegra sales look like they are going to do well, so as long as Fermi doesn't totally bomb, then they can probably have a pretty good year. Once stock brokers start freaking out about nVidia shares, then you should be worried.

On the other hand, Cracked.com taught me that throwing darts at a dart board gives you just as good of a chance at making money in the stock market as a stock broker does... so maybe not.
 
Ya know, Ill post it again, because you didnt bother to read my link. I usuallyu dont post links for nothing
Nvidia shares dip 9% on inventory buildup
http://sanjose.bizjournals.com/sanjose/stories/2010/02/15/daily82.html?ana=yfcpc
shares fell about 9 percent Thursday as investors apparently worried about an inventory buildup.

Santa Clara-based Nvidia (NASDAQ:NVDA) closed at $16.67, down 7 percent, after dipping as low as $16.21 during the day's trading.

On Wednesday the company reported fourth quarter net income $131.1 million, or 23 cents a share, compared to a loss in the same period a year earlier of $147.4 million, or 27 cents a share.

Revenue was $982.5 million, more than double the $481.1 million in the year-ago quarter.

The company also reported a 19 percent quarter-over-quarter inventory increase.



So, talking profits now, and taking them earlier, and renaming earlier, and not using GDDR5 earlier, and not going DX10.1 earlier and just kept making the same old cards, and renamed them, then renamed them some more has caught up with them.
 
There are some things that have escaped som of you, and I will try to explain what it is. For example a Phenom II C2 and a C3, The letter 'C' is about the base layer the masks etc. and the number '2 or 3' is about some material spin. The root to Nvidia's yield problems, are in the letter part, not in the number part. That's why yields will continue to be low.

“Nvidia’s chief executive officer did not provide any concrete timeframes for the transition of the whole lineup to the new Fermi architecture, but said that since the owners of mainstream and entry-level graphics cards hardly demand new functionality, it is not crucial for Nvidia to update currently available “fabulous” graphics chips. In addition, the speed of the transition depends on the supply of 40nm chips by TSMC.

All of that just depends on 40 nm supply and we are trying to finesse it the best we possibly can. For the entry-level products, the truth is that the new architectures […] are probably not extremely well appreciated anyhow. People, who buy the new architectures, tend to be early adopters and they tend to be the game enthusiasts, workstation designers or creative artists or – there are very specific reasons why it really enhances their experience. Our current-generation GPUs are fabulous and all the things that mainstream consumers would use their computer for. […] I think the mainstream GPUs are really fabulous and has been enhanced recently with some really great features and so my sense is that they are going to continue to do quite nicely in the marketplace. Then we will just transition as fast as we can,” said Mr. Huang.
Click to expand...

The ATI chips have extra via's and redesigned transistors, that's why they yield better, and can run higher frequences, at lower voltages. But it all must have cost a lot of money for ATI to do. But if they had not done it, they would have bin f***** to.



 


ATI's shares are also down, or at least there were as of the last quarter, as was nVidia. Intel stole it from both of them.

I will post again as well because you don't read mine exactly, nVidia is predicted to regain market share, nVidia's troubles right now mean nothing practically. They profited 3/4 quarters last year. They are at least double the size of ATI still, ATI hasn't gained any market share on them either.

But you're partially right, they have made a lot of mistakes, which can lead to bad things. But if Fermi does at least moderately well, nVidia will probably still have a better year than ATI financially. And I don't care how much nVidia's or ATI's street cred matters to you or anyone else, money is what speaks volumes, and nVidia simply has more of it.

nVidia at this point STILL isn't in trouble. Their Tegra sales are starting to pick-up as well. If Fermi does poorly, then and only then are they actually in trouble, right now they are just skirting the edge of it. And even if Fermi DOES do poorly, it really doesn't matter, cause they can float for a while on what they have.

It's tiring seeing doomsday comments all over these forums about nVidia going down when they aren't even close to doing so. It's almost like people WANT to see a monopolized consumer graphics market, which I do not understand for the life of me why; I want both ATI and nVidia to flourish, as should everyone else who doesn't want to have to start donating organs to get future graphics cards.
 
If you dug a little deeper with Nvidia you might have picked up on the fact that Nvidia stock trading volume is WAY up. IMO we probably are seeing maneuvering for the sharks to buy low , Fermi's launch is going to send the stock sky high and make some rich folks richer.
FEB 11,2010 http://www.marketwatch.com/story/tech-stocks-try-to-find-upward-momentum-early-2010-02-11
Graphics chipmaker Nvidia /quotes/comstock/15*!nvda/quotes/nls/nvda (NVDA 16.58, -0.09, -0.54%) rose 75 cents a share, or 4.6%, to $17.12 after Roth Capital analyst Arnab Chanda raised his rating on the company's stock to buy from hold. Chanda said Nvidia is likely to regain some of its PC business following the upcoming launch of its Fermi processor for PCs.
Click to expand...

Let me repeat this important part : Chanda raised his rating on the company's stock to buy from hold. Chanda said Nvidia is likely to regain some of its PC business following the upcoming launch of its Fermi processor for PCs
 
You are all missing important facts, and a few even denying them.
Someone said their inventories were shrinking, as theyve quit making parts, well, we know theyve quit making parts, so, that leads to larger inventories.
If these inventories are growing, and JHH said they didnt have enough 40nm, whats that telling you? And, since we all know that Fermis isnt coming anytime soon, whats that telling you?
Im not talking profits here, as Im sure the HPC market will pay dearly for Fermi.
Im talking about markets, and shares. The mobile markets are becoming more and more ATIs, and its the largest growing market as well.
If the high/mid end is dwindling simply because of EOL, then the inventory build up is from the low end/mid end, and again, when will Fermi products come?
If nVidia once again burns their partners and non saleable products at decent margins, and I think everyone needs to check to see what the 330 is, and how nVidia is using everything they can to call it that, to dump off product.
So again, if these things, (new fermi parts) dont start to happen til May/June in ernest, then theres simply no way the things being said can take place
 
I have read somwhere that they have 10K Fermi parts. If that is all they have out of 5K 5K$ wafers. Then they are expensive 😀
 
More from Charlie 'Douchebag' Demerjian the high priest of the Ati faithful, why am I not suprised. Don't you have any other sources? After all this is the guy who thought that the GT300 was Fermi even though it was actually the OEM 40nm shrinks and there are so many other things he has been wrong about why do you keep trying to put his rantings forth as gospel truth?
 
Mainly cause I agree with Apes assessment

True he's been hard on nV since the whole Bump-gate fiasco where they lied to him and then tried to discredit him, but with Fermi remember it's also usually been like this;


Fact <-----------x-----|---x----x------x-----> Fiction


.................Charlie.....Theo. nV....Fuad


Especially about things like yields and re-spins where Charlie would print it, Fuad would deride it, nV would deny it, and then weeks later the press statement would confirm it.

http://www.tomshardware.com/forum/page-281640_33_50.html

If youve been following it from day 1, it basically fits so far, unless all the sudden its changed
 
Jes, I did not initialy read his article, it sounded to 'Charlie going evil on Nvidea' at first :na: . But the supply can't be that great.
 
Im using his assessments so far, others as well, plus the underlying lateness/non response from nVidia itself.
Add those things up, and its even more "accurate". Ive heard from others that the 7xx clocks havnt been hit yet by Fermi, and thus the shader clock as well.
If you think about this, removing 15+% off the top of Perf because of the clocks, it makes sense doesnt it?
Ive been expecting around 20%, so to me it makes sense. It also makes sense the lessor card runs a tad faster, as weve seen in all previous iterations with less shaders.
If it runs this hot, ocing wont be done well, and we all know the 1Ghz 5870s are just right around the corner.
This all spells major problems for nVidia, even if perf is better than what Charlies saying, because say, 10% isnt enough really, Fermi needs that 20% gap
 
Nvidia need a halo product, end of story.

The R600 is a good representative of this. There was nothing really wrong with the chip, it wasn't as good as it could have been but the main problem was that it wasn't as fast as nvidia's offerings.
 
Heres a good reason to listen to Charlie:
If they have to do only one respin, that puts Nvidia out ten more weeks, eight for silicon, and two for testing. Nov 1 now becomes Jan 15. Since Nvidia is 'teh awsum' at making new parts of late, a second respin isn't all that much of a stretch, so it could be out around this time next year. We won't bring up why the GT212 was canned and the GT214 became the GT215, that would simply be impolite. We have no reason to badmouth TSMC like some greener folk who really should read DFM rules.
http://www.theinquirer.net/inquirer/news/1051818/nvidia-gt300-promised-october

So, it will be here this time from last year when he typed this, and no one else did, and everyone hated him and denied it etc
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom