Overclockaholics Forums - All things Fermi

Overclockaholics Forums (http://www.overclockaholics.com/forums/index.php)

- Hardware/Software News (http://www.overclockaholics.com/forums/forumdisplay.php?f=94)

- - All things Fermi (http://www.overclockaholics.com/forums/showthread.php?t=1561)

In detail NVIDIAs Rocket Sled Demo: DirectX 11, Physics, Fluids and More

Quote:

WITH ANOTHER LAUNCH of the Nvidia GT300 Fermi GF100 GTX480 upon us, it is time for an update on the status of that wayward part. Production parts have been coming back from TSMC for several weeks now, and the outlook for them is grim.

We first got word that production A3 GF100s were back in Santa Clara near the end of January. When enough of them rolled off the line to characterize the silicon, we hear there were no parties in Santa Clara. For those not interested in the 'why' side of things, the short answer is that the top bin as it stands now is about 600MHz for the half hot clock, and 1200MHz for the hot clock, and the initial top part will have 448 shaders. On top of that, the fab wafer yields are still in single digit percentages.

That said, the situation is far more nuanced than those three numbers suggest, and the atrocious yields are even after the chip has been downclocked and defective units fused off. To make matters even worse, the problems that caused these low yields are likely unfixable without a complete re-layout. Lets look at these problems one at a time.

Number one on Nvidia's hit list is yields. If you recall, we said that the yield on the first hot lot of Fermis that came back from TSMC was 7 good chips out of a total of 416 candidates, or a yield of less than 2 percent.

The problem that Nvidia faces can be traced to what it is doing to fix the issues they face. The three steppings of GF100 are all what are known as metal layer spins, something that is cheaper and faster than a full base layer respin, taking about two months to see results. A full base layer respin takes well over a quarter, likely more than six months to accomplish, and costs more than $1 million just for the masks. Metal layer spins are denoted by upping the number, A1 to A2 for example, while base layer respins up the letter, A3 to B1. Nvidia counts first silicon from A1, so the current A3 is the third metal spin.

Metal layer spins tend to solve logic problem like 1 + 1 = 3, not power or yield issues. Most yield problems are related to the process that the chips are made on, and modified by factors like how fast you try to run the transistors, how much you bend the design rules, and other related issues. While this is a very simplified version, metal layer spins don't tend to do much for power consumption or yield problems.

Full story at source

Source: Semiaccurate

Damn, sucks to be nVidia right about now. They should really start whoring out their gtx200's at half price to push for the funds for a base spin.... :Dizzy:

200s were extremely expensive to produce too I beleive, the poor price:performance ration was not so much them gouging the consumer just extremely expensive products to produce.

Quote:

Originally Posted by FACE411 (Post 26630)

Damn, sucks to be nVidia right about now. They should really start whoring out their gtx200's at half price to push for the funds for a base spin.... :Dizzy:

Until they do there market share will remain to be complete crap ...

260/216 should roll in the 150 range at least

285's I would say need to be around 250

You have to give it to ATI for getting the dx 11 out the door and even know they are struggling with drivers they have acomplishe their goal by stealing the high end with the 5870 and the midrange market with the 5850 witch is just a tick slower

NV will have to compete with a refresh but I see no reason why ATI would release this until Nv puts something out to compete with the current line up

I'm actually excited for anything new from nVidia. Ati has done a remarkable thing with the 5 series but all of this GSOD, driver stops and generally shit drivers has left me looking for something from the green team.

Quote:

Originally Posted by ReverendMaynard (Post 26684)

I couldnt agree more, although the 5870 is a fantastic card for getting benchmark scores, I would never in a million years use one or more of these things in my gaming rig.
Fermi will be in there as soon as it comes out (pending total suckage, which I cant imagine would happen) just because I know the drivers will work, SLI will work, and any issues that come up will be far easier to deal with than the damn ATI driver nonsense.

I hear ya. I have not had to deal with the CF issues since I only have one of them.

I sure hope when they do bring it out and it is a home run....

If not look for a lot of developmental people at nV to get the ax ........

You're over-exaggerating nVidia's current situation WD. What happens when AMD has to recall the 5 series? They can't fix it through driver revisions so they just have to admit that there is a hardware fault. OEM system manufacturers are applying pressure because the return rate is massive. People who pay $3000 for a pre-built rig don't bench it where the 5 series shines, they game on it and surf the internet, where the 5 series fails. Don't believe everyone who makes cliff notes from FUD articles brother, what's happening in the real world get's twisted by that bag of douche.

Quote:

Originally Posted by ReverendMaynard (Post 26690)

Could not agree more! I mean sure they have had their share of issues etc. but I'm sure once this thing comes out it will shine. ATI got the go on the DX11 stuff but as you see they raced so had and got a product out which has fault + the drives always has been shitty as far as I'm concerned. I guess we will see in a few weeks what is going to happen. :thumbsup: