• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

NVIDIA GT300 Yields Are Under 2%?

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Charlie over at SemiAccurate says that, according to what he has heard, yields of NVIDIA’s GT300 are currently at 1.7%. You read that right, under two percent. So, if what Charlie is reporting is even semi-accurate, yields are extremely bad.

How many worked out of the (4 x 104) 416 candidates? Try 7. Yes, Northwood was hopelessly optimistic - Nvidia got only 7 chips back. Let me repeat that, out of 416 tries, it got 7 'good' chips back from the fab. Oh how it must yearn for the low estimate of 20%, talk about botched execution. To save you from having to find a calculator, that is (7 / 416 = .01682), rounded up, 1.7% yield.
 
Like how I worked “semi-accurate” into that one? Heh…I’m good like that.
 
Steve, all of your articles are 'semi accurate'. :p

If Charlie is indeed telling the truth, then Nvidia is going to be screwed
 
Semi-accurates' more accurate than fudzilla, but this time I hope not - I want AMD 5800 series pricecuts fast!
 
Those yields are so low I simply cannot believe they are even close to semi-accurate! Here's hoping that this one is just FUD.
 
Semi-accurates' more accurate than fudzilla, but this time I hope not - I want AMD 5800 series pricecuts fast!

The difference between him and Fudzilla is that Fudzilla has actually been correct in the past. He hasn't.
 
Holy crap. That just don't sound right.
Below 2%???
I think I'm gonna call this one fud.
 
I don't think thats even possible unless there is some sort of major design flaw and/or they had a earthquake when they were making them.

If true, this is a knockout punch.
 
I'd believe 2% are bad...

but I won't believe nVidia screwed up this bad, especially TSMC has already worked out kinks during the 4770 line.

Unless if nVidia's chip design is redically different, or horrendously large, it isn't going to be that low.
 
I guess ATI's small chip strategy has some merit :p

Although if the R800 has 1600 steam processors, it must be a pretty big, complex chip as well. Calling a 2150 million transistor chip "small" doesn't seem right :eek:
 
sounds like they can still be profitable at their $400 launching price point to me...
 
That sounds like prototype yields to me. Obviously they need to increase before it can be sold at any reasonable level.
 
Well, if NVIDIA would have something to put against HD58x0, then they would not need to defend themself with PhysX propaganda - which they did. I think NVIDIA is going R600 route this time, they need to learn from their own mistake, they didn't learn from ATI mistake.
 
Well, if NVIDIA would have something to put against HD58x0, then they would not need to defend themself with PhysX propaganda - which they did. I think NVIDIA is going R600 route this time, they need to learn from their own mistake, they didn't learn from ATI mistake.

They already have.

I got a 'FX 5950".

Awesome. the first dual slot I had ever seen.
 
The difference between him and Fudzilla is that Fudzilla has actually been correct in the past. He hasn't.

You do know where Charlie used to be, where Faud used to be? So on and so forth? Both are from TheInq for years and a lot of what you read from TheInq that was accurate was written by Charlie. This is a rumor mill post, you know...he said - she said kinda thing.
 
um, the last rumor I heard was that it was under 30%, 1.7% is just a worthless wafer. They would not even be manufacturing at that point. I am going to have to call bullshit on this.


well maybe not, just charlie putting a very hard spin. he saying that this is from the "hot lot" or the pre manufacture side wafers. this does not mean that the production will be that bad, if its the case at all. for all we know they could have been using the initial 40nm process that had so many issues.
 
Was looking forward to price cuts in Octuber but guess i'll have to wait or go with the 4xxx line, eh.
 
Several experts in semiconductor engineering, some who have overseen similar chips, were asked a couple of loaded questions:

Loaded questions indeed. Reporters can and do put words into people's mouths.

Anyone remember when Charlie said that NVIDIA positively absolutely bet-your-life-on-it cannot put two GT200 cores out as a single graphics card? Uh yeah, that was swept under the rug pretty quick, as were his other inaccurate reportings.
 
Loaded questions indeed. Reporters can and do put words into people's mouths.

Anyone remember when Charlie said that NVIDIA positively absolutely bet-your-life-on-it cannot put two GT200 cores out as a single graphics card? Uh yeah, that was swept under the rug pretty quick, as were his other inaccurate reportings.
They didn't though. They had to wait for the 55nm shrink before they could.
 
Wow, that doesn't garner much confidence on a vid card that has that chip does it?
 
I can confirm this story. My FX5200 is in most cases slower than my TI4200
 
Yields for new designs are always very low. I don't see anything out of the ordinary here. We all knew Nvidia was behind ATI with the next gen. ATI always ships 3-6 months sooner than Nvidia. Then Nvidia ships a card that blows the doors of ATI's high end. What's new here?
 
Loaded questions indeed. Reporters can and do put words into people's mouths.

Anyone remember when Charlie said that NVIDIA positively absolutely bet-your-life-on-it cannot put two GT200 cores out as a single graphics card? Uh yeah, that was swept under the rug pretty quick, as were his other inaccurate reportings.

I remember him saying every G92 was bad..... still waiting for the sky to fall on that one.
 
ATI always ships 3-6 months sooner than Nvidia. Then Nvidia ships a card that blows the doors of ATI's high end. What's new here?

when was the last time ati shipped a card before nvidia?
 
FWIW, I heard very similar information from two other sources recently.
 
x1900 refresh compared to 7900

before that, only the 9700 series was actually ahead.

nvidia generally has always had a small lead on next gen, some more than others.
 
when was the last time ati shipped a card before nvidia?

I think he just means some generations/refreshed.

Sort of how the 'Top Dawg' GPU contest switches back and forth between nVidia and ATi.

Like how the 4870 was king, then the 295, now it's (about to be) the 5870...
 
I think he just means some generations/refreshed.

Sort of how the 'Top Dawg' GPU contest switches back and forth between nVidia and ATi.

Like how the 4870 was king, then the 295, now it's (about to be) the 5870...

yeah I know my times are off, but we get the point, right?
 
Back
Top