Apple says some MacBook Pros affected by faulty Nvidia chips

[Sun Baked]

sorry I don't get you

Apple karma with this sort of thing means turning in a machine with a perfect screen and needing repair for one with a bunch of stuck pixels, dents, and scratches.

Some of us think they put these returned machines aside for people known to be pain in the asses.

 

[dorenthe]

yeah I know...

 

[PDE]

I'm not the bad guy!!!
Fact is that I bought a laptop for 4000 A bucks and even if they replace the logic board THE WARRANTY for 2 months is gone because I couldn't do anything with the 9600 chip... I thought all the time it's a software problem...

that's what I say ... I bought a damaged ONE
...
The touchpad doesn't work in the beginning and I had some other issues...
also said Steve Job that the glossy display not reflect, haha.

The book was not able to work to 100 %.

The only thing what I want is new warranty after the logic board replacement or a new Macbook Pro ...

So I want to try the scratchie issue,
hey what's that a scratchie.
The macbook hasn't a scratchie before....
oh I WANT A NEW ONE

Of course it is not ethical to use the scratch you made to try to get a replacement. On the other hand, Apple is not handling this well and its behaviour in connection with the Nvidia chip is unethical too. YOu're right, you bought a defective computer (and Apple knew it then too!) and you should get a new replacement. Everybody SHOULD. The problem is that Apple has already chosen not to do a recall and they're not going to change their mind.

If I were you, I'd write a letter to the BBB and Attorney General in your state, clearly, dispassionately and concisely detailing what happened, together with a clear request, i.e. a replacement. Apple executive customer service will contact you and you can take it from there. You can also try writing sjobs@apple.com -- executive CS usually responds to mail sent there too and that might be faster/easier.

Good luck!

 

[dorenthe]

thx buddy
I'll try it

 

[MowingDevil]

I'm not the bad guy!!!
Fact is that I bought a laptop for 4000 A bucks and even if they replace the logic board THE WARRANTY for 2 months is gone because I couldn't do anything with the 9600 chip... I thought all the time it's a software problem...

that's what I say ... I bought a damaged ONE
...
The touchpad doesn't work in the beginning and I had some other issues...
also said Steve Job that the glossy display not reflect, haha.

The book was not able to work to 100 %.

The only thing what I want is new warranty after the logic board replacement or a new Macbook Pro ...

So I want to try the scratchie issue,
hey what's that a scratchie.
The macbook hasn't a scratchie before....
oh I WANT A NEW ONE

If you aren't the "bad guy" then be honest.
*deep breath* I can only hope you're not an adult with such fuzzy morales.

Going through issues like the nVidia debacle isn't easy on anyone but it doesn't justify fraud.

 

[jjahshik32]

A chemist's theory on why the latest news is ok for NVidia:

First, two references:
[1]http://www.theinquirer.net/inquirer/news/947/1013947/nvidia-should-defective-chips
[2]http://www.dvhardware.net/article31847.html

In [1], Charlie discusses how the NVidia chip problem has two parts: the bumps and the underfill. An underfill with a higher glass transition temperature protects the bumps better. This is an old article from when the problem was first discovered. The whole original 3-part series is good reading.

In [2], NVidia discusses how the underfill was changed for these new chips. My impression is that the info quoted here was also given to Charlie, but he chose not to present it.
quote:
Hara talked about how the original problem announced by Nvidia on July 2 was rectified. "A more robust underfill would have taken the stress off the bumps and kept that (original problem) from happening. What we did was, we just simply went to a more robust underfill. Stopped using that (previous) underfill, kept using high-lead bumps, but we changed the underfill. And now we don't see the problem."


In the latest article (about the 9600M and the new MBPs), they only do an analysis of the bumps. While the bumps remain high-Pb, the underfill is not mentioned -- and is presumably changed, as Hara asserts. Charlie completely ignores the possibility of a newer underfill mitigating the problem of Pb bumps. If the underfill is changed, it is quite possible that the failure rate problem of the parts has been corrected. Only a true chip packaging / materials engineer with access to failure rate data for that specific underfill / bump combo would know for sure. But, if NVidia is talking in public about the issue, and commenting on specific technical changes (underfill), I'm willing to give them the benefit of the doubt for the 9600 part under consideration.

That said, there may be other design errors which lead to failure problems for these parts unrelated to the packaging -- just like with any new CPU/GPU. Just be careful about what's being blamed as the culprit.

Disclaimer: I own a late 2007 MBP with the 8600M which has the "known bad" packaging, and I use it to play games multiple times a week, and it has yet to fail (knock on wood).

found this from someone's post in arstechnica. So this pretty much means Nvidia has fixed the problems in the 9600m gt.

 

[madmaxmedia]

found this from someone's post in arstechnica. So this pretty much means Nvidia has fixed the problems in the 9600m gt.

I was going to thank you for doing all that research and typing up that explanation, until I got to the last line.

But thanks for posting it. IIRC the writer asked Nvidia who could've elaborated on why the new chips weren't a problem, but chose not to. I'm not sure why, since it appears that hopefully all is well in Unibody-Land.

 

[JavaWizKid]

Does that mean the 8600m that Apple are replacing broken ones with have changed underfill to stop the high failure rate problem?

 

[pionata]

I'm starting to have problems with the video card, 8600m.

A friend of mine had the screen go totally blank.

10 days left to get my apple care...

I guess the crappier their computer gets, the more apple care package they sell.

Worse mac built quality I've ever seen, amazing machine but poor pieces.

 

[Beau10]

found this from someone's post in arstechnica. So this pretty much means Nvidia has fixed the problems in the 9600m gt.

Uh huh... how about I quote you from last week?

http://forum.notebookreview.com/showthread.php?t=325301&page=9

"Hmm.. I found something interesting: http://www.theinquirer.net/inquirer/...ledgebase-g92s


"NVIDIA will transition from using high-lead solder (95%Pb/5%Sn) to eutectic solder (63%Sn/37%Pb) flip-chip bump material for the G92 product family. During the transition period NVIDIA will be supplying both high-lead and eutectic bump until inventory is depleted. No other materials are being changed."

So it means its still a mixed bag of the early released batches of the unibody mbp. Maybe this is why that some people are getting the black screen while minutes of gaming and others are not (including me)?"

 

[madmaxmedia]

Uh huh... how about I quote you from last week?

http://forum.notebookreview.com/showthread.php?t=325301&page=9

"Hmm.. I found something interesting: http://www.theinquirer.net/inquirer/...ledgebase-g92s


"NVIDIA will transition from using high-lead solder (95%Pb/5%Sn) to eutectic solder (63%Sn/37%Pb) flip-chip bump material for the G92 product family. During the transition period NVIDIA will be supplying both high-lead and eutectic bump until inventory is depleted. No other materials are being changed."

So it means its still a mixed bag of the early released batches of the unibody mbp. Maybe this is why that some people are getting the black screen while minutes of gaming and others are not (including me)?"

What the quoted text is saying is that it doesn't matter what the composition of the bumps is, if the underfill is robust enough. If that is the case, then it may not matter whether the bump material is high-lead or eutectic.

That scenario would make more sense from Nvidia's perspective. They may have blown it once before, but I doubt they'd willingly repeat the same mistake, which is basically only going to invite more bad press, scrutinity, and repair/replacemnt expense.

 

[Beau10]

What the quoted text is saying is that it doesn't matter what the composition of the bumps is, if the underfill is robust enough.

Certainly not the text I quoted, otherwise that's an amazing stretch -

"NVIDIA will transition from using high-lead solder (95%Pb/5%Sn) to eutectic solder (63%Sn/37%Pb) flip-chip bump material for the G92 product family."

Pretty darn literal. No room for misinterpretation. NVidia is quoted here as saying they are changing bump material, period.

Now the G92 is a desktop chip, but one with the alleged issue.

They may have blown it once before, but I doubt they'd willingly repeat the same mistake, which is basically only going to invite more bad press, scrutinity, and repair/replacemnt expense.

My belief is what happened in this case is they changed both bump and underfill. But the underfill was an easier/quicker process to remedy, the bump was more of a production line rework, and so in essence the former probably is true and the latter is being phased in.

 

[Habakuk]

I am afraid my MBP (please see sig) is the next candidate. It worked flawlessly until yesterday. I was watching a YouTube video and all of a sudden the screen was scrambled (distorted/inverted graphics and video). The animated scrambling show lasted about three seconds then the screen freezed. No more input possible.

After a forced shutdown and reboot OS X came up once again with distorted video and freezing after three minutes. The next restarts just came to the start sound and Apple logo - ten seconds later the grey curtain and the international text "Restart again...". Kernel panic.

Now I was at the local Apple Service Provider (Margaretenstraße) and at first they wanted 84 Euro for any estimates. I'll get that back if the damage falls under warranty. But they couldn't tell anything. "Maybe it's the video chip - maybe not - maybe there is an extended warranty - maybe not. We will have a look at it." I should hear from them after weekend.

(Typing on iPhone...)

Edit: They have called back two hours ago - earlier than expected. "The motherboard is defect and it will be changed under warranty." (Look in my signature when I've bought it.) - "Fine. Do you mean it's the graphic chip what is defect?" - "Maybe. The tech guy didn't explain it exactly."

So far - so good. On monday or tuesday it will be ready.

 

[madmaxmedia]

Certainly not the text I quoted, otherwise that's an amazing stretch -

"NVIDIA will transition from using high-lead solder (95%Pb/5%Sn) to eutectic solder (63%Sn/37%Pb) flip-chip bump material for the G92 product family."

Pretty darn literal. No room for misinterpretation. NVidia is quoted here as saying they are changing bump material, period.

Now the G92 is a desktop chip, but one with the alleged issue.

Oh, I understand that one. But what the previous post seems to indicate is that prior to this, this changed the underfill so that the GPU would be okay regardless of the bump material.

My belief is what happened in this case is they changed both bump and underfill. But the underfill was an easier/quicker process to remedy, the bump was more of a production line rework, and so in essence the former probably is true and the latter is being phased in.

Yeah, that's exactly what I meant. If that indeed is what happened, then I think unibody MBP owners have nothing to worry about.

 

[jjahshik32]

I am afraid my MBP (please see sig) is the next candidate. It worked flawlessly until yesterday. I was watching a YouTube video and all of a sudden the screen was scrambled (distorted/inverted graphics and video). The animated scrambling show lasted about three seconds then the screen freezed. No more input possible.

After a forced shutdown and reboot OS X came up once again with distorted video and freezing after three minutes. The next restarts just came to the start sound and Apple logo - ten seconds later the grey curtain and the international text "Restart again...". Kernel panic.

Now I was at the local Apple Service Provider (Margaretenstraße) and at first they wanted 84 Euro for any estimates. I'll get that back if the damage falls under warranty. But they couldn't tell anything. "Maybe it's the video chip - maybe not - maybe there is an extended warranty - maybe not. We will have a look at it." I should hear from them after weekend.

(Typing on iPhone...)

Edit: They have called back two hours ago - earlier than expected. "The motherboard is defect and it will be changed under warranty." (Look in my signature when I've bought it.) - "Fine. Do you mean it's the graphic chip what is defect?" - "Maybe. The tech guy didn't explain it exactly."

So far - so good. On monday or tuesday it will be ready.

I noticed that the most problems that people are having with is the last gen 15.4" mbp with the 2.2ghz and 128mb vram machines. So it doesnt really surprise me.

 

[sclough]

I noticed that the most problems that people are having with is the last gen 15.4" mbp with the 2.2ghz and 128mb vram machines. So it doesnt really surprise me.

I would assume that's just because those machines are a little older. Are you suggesting the 128mb machines are more prone to fail than the later 256mb machines?

 

[cohenben]

I think extending the 2-year warranty was a very fair move by Apple.

Most people I've consulted on this issue will confirm that a faulty graphics card will fail within a year (or two) of using it. With apple care you get an additional year on top of that (3-years) so you're pretty well covered if you have one of these failing cards.

I don't know when they introduced the 8600 card into the MBPs, but has anyone experienced a failed card beyond 2 years of use? Or is it too early to know this?

 

[jjahshik32]

I would assume that's just because those machines are a little older. Are you suggesting the 128mb machines are more prone to fail than the later 256mb machines?

No because the 2.4ghz mbp (the higher end back then) doesnt seem to still have too much problems as the 2.2ghz mbp.

Even 10 months ago when the problem first came up, most people from several different threads that complained about their mbp having gpu issues or just plain wont show anything on the screen mostly had the 2.2ghz mbp with the 128mb vram gpu.

 

[PDE]

I think extending the 2-year warranty was a very fair move by Apple.

Most people I've consulted on this issue will confirm that a faulty graphics card will fail within a year (or two) of using it. With apple care you get an additional year on top of that (3-years) so you're pretty well covered if you have one of these failing cards.

I don't know when they introduced the 8600 card into the MBPs, but has anyone experienced a failed card beyond 2 years of use? Or is it too early to know this?

Except that they didn't extend it by two years -- they extended it from one year to two years. Given that these machines might fail prematurely (in my book that means within 5 years, not 2 years), I think it is crazy and terribly unreasonable of Apple. It's Stingy and also disrespectful to its customers. If it had been any other company I would have vowed to never by another product from them. Unfortunately, I love os x and that means putting up with this kind of BS.

Remember, we're not talking about a battery or something that is cheap to fix, so once it fails the computer will need to be repaired and, if out of the 2-year windown, it will be very expensive.

 

[JavaWizKid]

I'm sure the issue has been fixed by now. Since the 9600s use high lead, but have a better underfill, it overcomes the fault. So if Nvidia did this to the 8600s, they would fix the issue and wouldn't need a structure redesign to suit eutectic solder. If this is true, Apple didn't need to extend the warranty past 2 years because:
1. Apple Care would be useless buying unless you have other problems.
2. When your MBP dies which probably will be within the 2 year timeframe, it will be fixed with a better underfill one which wont suffer from the same issue.

Correct me if I'm wrong.

 

[PDE]

I'm sure the issue has been fixed by now. Since the 9600s use high lead, but have a better underfill, it overcomes the fault. So if Nvidia did this to the 8600s, they would fix the issue and wouldn't need a structure redesign to suit eutectic solder. If this is true, Apple didn't need to extend the warranty past 2 years because:
1. Apple Care would be useless buying unless you have other problems.
2. When your MBP dies which probably will be within the 2 year timeframe, it will be fixed with a better underfill one which wont suffer from the same issue.

Correct me if I'm wrong.

You're wrong because of your assumption that it will fail within 2 years. We don't know that yet since the chip has not been used for that long. I doubt anybody can predict when it will fail, even if it may be more likely to fail earlier. Customers should be able to rest assured that whenever it fails, Apple will fix it for free. Even if that is three years down the line. My family and I have had many laptops over the years and never once experienced a GPU failure, even with computers that are 7-8 years old.

Apple should extend coverage of the GPU for as long as they support the product with repairs, anything shorter is unfair to the customer.

 

[JavaWizKid]

The trend seems to be that just after the standard warranty the 8600 fails. Every report of its failure is within the 2 year span so I am very comfortable with Apple's decision. Also, why do you think it should be covered for whenever it fails? The absolute maximum warranty you can get is 3 years anyway. So if your chip fails after 3 years would you complain?

What I meant by "Correct me if I'm wrong" was whether the underfill has been changed or not in the new 8600s.

 

[PDE]

The trend seems to be that just after the standard warranty the 8600 fails. Every report of its failure is within the 2 year span so I am very comfortable with Apple's decision. Also, why do you think it should be covered for whenever it fails? The absolute maximum warranty you can get is 3 years anyway. So if your chip fails after 3 years would you complain?

What I meant by "Correct me if I'm wrong" was whether the underfill has been changed or not in the new 8600s.

Well, obviously, we haven't gotten to two years yet, so we really don't know what will happen after that. Just because many people have experienced this just after a year doesn't meant that there won't be even more people experiencing it after two years. GPUs shouldn't fail for many many years, long beyond the productive life of the computer as a whole. Apple extended the warranty coverage for the logic boards of the Powerbook 5300/190 series to 7 years because of various defects -- that's reasonable. If the GPU fails after three years and everything else is working perfectly, why should I have to pay for it when I know that it's because of an inherent fault that was there when I bought the machine? This is not a warranty issue. It's a production defect that should be fixed by Apple whenever it causes trouble, even if it's outside of warranty.

I don't quite understand how anybody could be happy with a one year extension when the defect is so serious. We paid a lot for these machines. If they really only fail within the first two years of use, then Apple shouldn't worry about extending it for 3-4 years since there would be no claims anyway. Above all, extending the coverage of the GPU defect for the normal life of the product is about goodwill and restoring confidence in their products. Apple and Nvidia ****ed up with this one (mostly Nvidia), and customers should not shoulder ANY of the cost of that beyond the already considerable inconvenience of having to send it in for repair.

 

[sclough]

No because the 2.4ghz mbp (the higher end back then) doesnt seem to still have too much problems as the 2.2ghz mbp.

Even 10 months ago when the problem first came up, most people from several different threads that complained about their mbp having gpu issues or just plain wont show anything on the screen mostly had the 2.2ghz mbp with the 128mb vram gpu.

Is there anything technically different between the two that would cause the 2.2ghz to fail more than the 2.4ghz? The gpu's are identical so shouldn't they be soldered to the logic board in the same way or am I missing something?

 

[cohenben]

Well, obviously, we haven't gotten to two years yet, so we really don't know what will happen after that. Just because many people have experienced this just after a year doesn't meant that there won't be even more people experiencing it after two years. GPUs shouldn't fail for many many years, long beyond the productive life of the computer as a whole. Apple extended the warranty coverage for the logic boards of the Powerbook 5300/190 series to 7 years because of various defects -- that's reasonable. If the GPU fails after three years and everything else is working perfectly, why should I have to pay for it when I know that it's because of an inherent fault that was there when I bought the machine? This is not a warranty issue. It's a production defect that should be fixed by Apple whenever it causes trouble, even if it's outside of warranty.

I don't quite understand how anybody could be happy with a one year extension when the defect is so serious. We paid a lot for these machines. If they really only fail within the first two years of use, then Apple shouldn't worry about extending it for 3-4 years since there would be no claims anyway. Above all, extending the coverage of the GPU defect for the normal life of the product is about goodwill and restoring confidence in their products. Apple and Nvidia ****ed up with this one (mostly Nvidia), and customers should not shoulder ANY of the cost of that beyond the already considerable inconvenience of having to send it in for repair.

but it's a computer and of all manufacturing feats of the last 100 years you'd be hard pressed to find one with more part failures across the board. Even cars are more reliable than computers.

My point is that you have the option for that three year warranty with apple care and if you continue to use your machine past three years consider yourself lucky until your computer suffers it's ultimate and unavoidable failure...whether that is in the form of the GPU, hard drive, cooling fan or spacely sprocket.

I agree that if these nvidia 8600 chips have a major meltdown at 3.5 years that apple and/or nvidia should offer some form of remedy. All indication is that apple will take such a high road because when learning that some cards will unfairly fail at just over one year they quickly extended their warranty to two years.

Consumers deserve to be protected and I'll be the first to organize a class action suit if apple were to act unfairly. But umtil there is evidence that chips will fail after two years at the fault or negligence of the manufacturer (apple OR nvidia) then I give full credit to apple's current extention of their standard warranty.