AMD: R9xx Speculation

[OgrEGT]

Well, isn't it possible, that at a certain power draw (which was not originally intended) which would be needed to match up (or overtake) the GTX580's performance revealed, that this MOSFET part, which was in question, was not being adequate?

If one would decide, to communicate to manufacturing (all launch cards are built by AMD) to replace this part with another one, would that be manageable in a few weeks? (Depending on the amount of already built cards)

 

[3dilettante]

An unintended power draw sounds like a failure in qualifying the design, so "fixing" the problem with a different component means a successful qualification run is still needed on the fixed design.

 

[neliz]

Well, isn't it possible, that at a certain power draw (which was not originally intended) which would be needed to match up (or overtake) the GTX580's performance revealed, that this MOSFET part, which was in question, was not being adequate?

It's not a MOSFET, but a DrMOS and if they can't supply enough power in a 7+1 setup.. there's no enough PCB space to do it with traditional MOSFETs

 

[Gipsel]

[..] that this MOSFET part, which was in question, was not being adequate?

If one would decide, to communicate to manufacturing (all launch cards are built by AMD) to replace this part with another one, would that be manageable in a few weeks? (Depending on the amount of already built cards)

I've just read somewhere that the lead time for ordering larger quantaties of those parts is supposed to be something like a full quarter or even more. Sounds to be quite a bit on the high side if you ask me, or it is some quite new and special stuff

 

[Mize]

I've just read somewhere that the lead time for ordering larger quantaties of those parts is supposed to be something like a full quarter or even more. Sounds to be quite a bit on the high side if you ask me, or it is some quite new and special stuff

Not in today's JIT marketplace. I have a few chips used in embedded products that are SIX MONTH lead time right now because the foundries down-scaled for the recession.

 

[LordEC911]

Not in today's JIT marketplace. I have a few chips used in embedded products that are SIX MONTH lead time right now because the foundries down-scaled for the recession.

Yep, many fabs laid off a lot of very skilled employees when things hit bottom to quickly cut expenses. Unfortunately, once things start to rebound they need to not only find qualified employees but also train them which can take quite awhile.

I know someone at Microchip and they were simply making employees take time off until things got better, so that they could ramp up as quick as they needed when things got better.

 

[nyt]

From the GF100 thread:

690/1380/3800 MHz​

Market release
12/8/2010

http://bbs.expreview.com/viewthread.php?tid=37829&from=recommend_f

Cayman delay caused by this, IMO.

 

[SimBy]

From the GF100 thread:



Cayman delay caused by this, IMO.

Doubt it.

 

[thatdude90210]

Doubt it.

You don't believe that a photoshopped gtx 580 picture combined with approx. gtx 480 specs caused Cayman's delay?

 

[Razor1]

Don't know if this was shown before......

 

[Gipsel]

Don't know if this was shown before......

That picture doesn't show any knowledge about Cayman, it is just a combination of known data and the expectations of the author.

 

[OgrEGT]

That picture doesn't show any knowledge about Cayman, it is just a combination of known data and the expectations of the author.

Wouldn't it be at least a (further?) confirmation for the chipsize? From the picture it seems that Cayman is indeed <400mm2. More like 360-370mm2.

 

[AlphaWolf]

Wouldn't it be at least a (further?) confirmation for the chipsize? From the picture it seems that Cayman is indeed <400mm2. More like 360-370mm2.

Would be if it were some leaked official AMD document. It's not.

 

[OgrEGT]

Would be if it were some leaked official AMD document. It's not.

Yes maybe you're right. The author of the diagram could have placed the info for Cayman on the basis of either his own speculation, or on the basis of true knowledge. I cannot say either the one or the other...

 

[nyt]

You don't believe that a photoshopped gtx 580 picture combined with approx. gtx 480 specs caused Cayman's delay?

Feel free to nitpick, but the info behind this is a GTX570 is coming which may grab a slice of the pie targetted by AMD.

 

[BRiT]

Cayman delay caused by this, IMO.

Nope. That's not the cause.

 

[caveman-jim]

Nope. That's not the cause.

What is it, then?

 

[SimBy]

What is it, then?

Supply issues?

You honestly think AMD would delay a launch cause of nV?

 

[caveman-jim]

You honestly think AMD would delay a launch cause of nV?

Nope.

 

[nyt]

Supply issues?

You honestly think AMD would delay a launch cause of nV?

Like your schedule isn't affected by what your competition is doing I believe they can afford to wait a bit and tweak a few things so the match against 570 is won without a doubt. And by the way, nobody but a few know for sure, so what you and I believe doesn't matter much in the end