Q: 2011 MacBook Pro and Discrete Graphics Card

D3us wrote:

Other possibilties are "head in pillow", can break later due mecanical stress too.
Tin wiskers, as mentioned, but less likely in this case.

You look well informed about the subject; why do you tin-whiskers would be less probable? did you have past experience with them?

CindyBruce wrote:

Imagine for a moment that Apple actually did not know what the problem was these past few years.  Would the situation be any different than what you see in these many threads and forums?

I think they did not know the problem. Many soldering artifacts related to BGA and lead-free had come up these years because of the largest audience and productive process. Apple's business is mainly based on customer's trust and loyalty. Knowledge is son of failure. We're evolving :)

carl wolf wrote:

I'm not sure that I can enlight you, but the totality of the evidence presented by users - and several referenced documents - indicate otherwise.  Apple vendors have attached several billion electronic devices, using a lead-free soldering process.  Apparently, only several thousand graphics chips are affected. Please, show me how you came to that conclusion.

     

Yes on three different models:
macbook pro 2010
macbook pro 2011
iMac 2011
all on GPUs. I don't think they are several thousand otherwise the other 2 models won't be recalled.

carl wolf wrote:

You - and others - are now convinced that the problem is related to tin whiskers.

I'm the only one who told that this could possibly be the problem or part of it and worth some investigation. This before i found this:
http://blogs.indium.com/blog/lead-free-silver-solder/page/2 which lead us to this(!):
http://www.dbicorporation.com/rohsbib.htm
where I discovered that there are plenty of likable friends (a.k.a. soldering artifacts) which more or less could suit our issue!:
Aristotle wrote about tin pest about 350 B.C.
H. Y. Hunsicker wrote about tin whiskers in 1947.
D. J. Lando wrote about conductive anodic filaments in 1979.
L. R. Conrad wrote about creep corrosion in 1982.
Elke Zakel wrote about Kirkendall voids in 1992.
Howard D. Blair wrote about copper dissolution in 1994.
Jason Fullerton write about head in pillow in 2008.
Mudasir Ahmad wrote about pad cratering in 2008.
Masato Shimamura wrote about non-wet opens in 2009.
Paul Reid wrote about foil cracks in 2010.
Werner Engelmaier wrote about trace buckling in 2010.
As you can see the last 5 are quite recent and I've discovered they're related to lead-free and BGA:

head in pillow
http://www.aimsolder.com/sites/default/files/head-in-pillow_bga_defects.pdf
"Head-in-pillow defects have become more prevalent since BGA components have been converted to lead-free alloys"

Pad cratering
http://www.ipc.org/feature-article.aspx?aid=Tests-for-pad-cratering
"Pad cratering is a growing issue as lead- and halogen-free assemblies grow. IPC-9708 provides testing techniques that eliminate the problem."

non-wet opens
http://hdpug.org/sites/all/files/Project_Pages/Package%20Warpage/SMTAI2012_MFX4. 3_Amir.pdf
"Also, environmentally friendly manufacturing is driving the use of Lead Free, RoHS Compliant SMT material and processes, which increased the processing temperatures of the component and board assemblies. All of these trends increase the challenges of board assembly. One of these challenges is the reduction in solder joint yields due to the increase in the occurrence of SMT Non Wet Open (NWO) solder joint defects."

foil cracks
http://books.google.it/books?id=dOFYjXIDm9YC&pg=PA14&lpg=PA14&dq=modes:+barrel+c racking+and+foil+cracking+(particularly+on+1%E2%81%842+oz+copper+foil)&source=bl &ots=jj500bRDFs&sig=hCKZx1lvS0calrp6aOxCWMnEw88&hl=it&sa=X&ei=5zDIU_OkBcfT7AbxtY H4CA&ved=0CCIQ6AEwAA#v=onepage&q=modes%3A%20barrel%20cracking%20and%20foil%20cra cking%20(particularly%20on%201%E2%81%842%20oz%20copper%20foil)&f=false
"PTH (pin through hole) reliability was substantially reduced by lead-free soldering due to two modes: barrel cracking and foil cracking (particularly on 1/2 oz copper foil)"

trace buckling
http://wenku.baidu.com/view/ce8f8c48e45c3b3567ec8be7.html
"I certainly have been railing against the RoHS ban of Pb in solders ever since I first heard about it[...]"

and our favourite
tin whiskers
http://nepp.nasa.gov/whisker/reference/tech_papers/2011-kostic-Pb-free.pdf
http://www.nbcnews.com/id/21151552/ns/technology_and_science-science/t/danger-le ad-free-electronics-tin-whiskers/
http://www.calce.umd.edu/lead-free/tin-whiskers/
http://www.lockheedmartin.com/us/news/features/2013/tin-whiskers.html
http://www.theguardian.com/technology/2008/apr/03/research.engineering

now you know how I came to the conclusion it's somewhat of the above and why, to me, seems the issue is related to lead free + BGA

carl wolf wrote:

Most users have had failures after years of service, and others only after a few hours...and you believe that it's caused by the growth of tin whiskers.  Please enlighten me.

You asked my opinion. this is my opinion: Apple is simpy reflowing old MOBOs with same problem. that's why they fail shortly. they don't fail? new MOBO's or MOBO's which was refurbished for other problems. many MBPs never failed? less use/they'll fall shortly/luck with good rare soldering

carl wolf wrote:

According to at least one of the referenced articles, "mechanical-loading" of a device attached using a lead-free soldering process increases the incidence of the growth of tin whiskers.  I'd like you to post your theory why this would occur, and I'll see if it meets with reality.  Regardless, in the MBP, more than one device is mechanically-loaded yet, only the graphics chip is affected.  Please, tell me why.

heating issue on GPU:
ps3
a sony vaio of a colleague of mine
MBP 2010/2011
iMac 2010
looks to me only GPU keep failing for heating and soldering. In fact you can reball it. If the problem would rather be in the chip you couldn't. That said my opinion is that GPU gets more heating warming cycles than CPU (which is soldered and dissipate heat in the same way). that is the only difference to me.
If you're asking why tin whisker grow: nobody know! I can just guess: they're crystals so they need three ingredients:
a supersaturation of the solute (nanoparticles of tin) in the solvent (air) (let's omit I don't like dividing matters in states but I rather prefer density)
Atomization
Crystal growth
tin particles fall and aggregate near electro-magnetic fields and keep growing. Why would lead avoid (or better: helps to stem) the problem? Imagine arabic bread and baguette. Bend them. check how many breadcrumbs are fallen by one and the other. well: leaded is arabic bread and lead-free is baguette. not to mention lead is diamagnetic (insusceptible to magnetic field)

1. the only reason I would exclude tin whisker is that reflowing would melt them and problem should be gone forever, but instead it comes back. but what if tin-whiskers damaged the GPU?
2. to spot soldering artifacts you can't just remove GPU by heating it or all the artifacts would melt. they can only rip it off with a pair of pliers (please don't do it: it would quite certainly avoid your warranty and any chance for a recall :D) and check with an electronic microscope pin by pin artifact by artifact
3. probably they'll check interrupting pin by pin to see if they can "simulate" the issue
4. the problem came up in 3 different lines in 2 years. let's see if they're lucky and next year we won't see the 2012 line starts giving the same troubles.
5. I want to know the problem because I heard people saying "no more apple". If it's what I think, I'd rather start asking myself how much my next computer would cost daily (price/(365*2)) (2 are the years of warranty in EU)
6. if this is the problem, a reballing laboratory would be one of the most profitable business in the next 5 years (then it should became cheaper so less profitable thank to the competition)

I hope I helped someone to learn something new, and I hope to learn something from you back.

I have a macbook pro 17" early 2011  last Sep the machine AMD GPU died and got it repaired less than a year later its dead again another 1500 dollars in parts and labour  these uni-body macbooks have a serious design flaw  they will all fail in the same way  the product should of been recalled.

S

Hi, I read on another thread, that someone got their logic board fixed and apple paid for it. They also said it was a quality control problem and apple would be announcing it soon. I don't know if it's true, but the link is here.

loloscott78 is the poster.

Re: problems with amd gpu on early 2011 macbook pro

At least Apple store locations should have been posted...

Well,  I really hope this is true and recall is starting next week or month!

"did you have past experience with them?"

Yes, and that's the difference between us.  You've read about tin whiskers, and I've dealt with them.

"If you're asking why tin whisker grow: nobody know!"

As a matter of fact, those involved in fabrication technology do understand the situation.  There's a difference between understanding the problem, and developing the solution.  Five hundred years before we arrived on the moon, mathematicians knew what was necessary to do so.  However, the technology did not support that endeavor.  You may believe that technology develops on a linear path, but it does not.  The first time that luggage appeared with wheels was 3 years after the US landed on the moon.

"Imagine arabic bread and baguette"

No, thanks.  The effect of mechanical loading on the growth of tin whiskers is not analogous to bread crumbing.  Regardless, there seems to be some confusion regarding the division of labor.  It's not up to you, or other users, to solve this problem.  That's squarely within the realm of Apple and its vendors.  It's up to end users to report the failures to Apple directly, and not via a discussion thread.

"did you have past experience with them?"

Yes, and that's the difference between us.  You've read about tin whiskers, and I've dealt with them.

"If you're asking why tin whisker grow: nobody know!"

As a matter of fact, those involved in fabrication technology do understand the situation.  There's a difference between understanding the problem, and developing the solution.  Five hundred years before we arrived on the moon, mathematicians knew what was necessary to do so.  However, the technology did not support that endeavor.  You may believe that technology develops on a linear path, but it does not.  The first time that luggage appeared with wheels was 3 years after the US landed on the moon.

"Imagine arabic bread and baguette"

No, thanks.  The effect of mechanical loading on the growth of tin whiskers is not analogous to bread crumbing.  Regardless, there seems to be some confusion regarding the division of labor.  It's not up to you, or other users, to solve this problem.  That's squarely within the realm of Apple and its vendors.  It's up to end users to report the failures to Apple directly, and not via a discussion thread.

I have an early 2011 mbb i7 2.3GHz QM with AMD 6750M and Intell 3000 HD and had the same problem and it appears to be 100% corrected.

Not long ago, over about a 5 week period, my early 2011 mbp's performance started nosediving, then the split screen, other screens, no login prompt, blue screen, etc.  Then there was zapping the PRAM, resetting the SMC, etc. sometimes with success and sometimes without but even the successful boots were short-lived.  Now prior to that 5 week period, for months I was having performance issues have to force quit processes, etc.

The last few days of it going south were the worst and the last day no matter what I tried I could not get the mbp to boot up.  Then came the fix which had nothing directly to do with Apple's claims.

I have to assume the fix was proper because I've yet to encounter a single issue with any current performance issues being almost negligible, the cooling fans rarely turn on and when or if they do, it's always low rpm and quiet.  I've watched long and short videos, burned a few high-rez 192/24 cd's while listening to an iTunes podcast, written emails, etc and I've not once encountered another freeze or bad screen of any sort.  And though I've never consider this mbp's performance as stellar, the performance today is perhaps as stellar as it's ever been.  More importantly, my confidence in my mbp has been restored in full as I'm typing on it now with numerous windows and browser pages open and Outlook open without any fear whatsoever of a split or frozen screen and frozen mbp.

I have over 30 years in the IT profession, mostly Oracle and Unix, and though I've never claimed to be the sharpest tool in any shed, I can confidently say what the problem is not, where the problem is, and that a proper fix should cost pennies on the dollar when compared to the cost of replacing or even re-balling a logic board.

This is not to say that the problem won't reoccur someday.  I certainly cannot guarantee that.  But even if it dropped dead tonight, that should not change where the problem isn't and where it most likely is and I should still be able to demonstrate these matters with the same confidence.

Although I'm always willing to help my fellow man, I see no reason why I should give this away freely, especially when so much is at stake. Hence, I emailed Apple to offer a proposal that if I could demonstrate what the problem isn't and what the problem is so that Apple's technicians could zero in quickly on the exact spot, he would honor my request.

Recently I was contacted by Corporate Executive Relations Jessica H. who thought I should freely provide what information I have.  The call ended with Jessica saying, if you change your mind, please contact me and I said likewise.  I then followed up with one more email (which Jessica will probably read first) just to update him and remind him ever so respectfully that we live in a capitalist society, not a socialist society.

Anyway, if they should change their minds I'm confident I can fulfill my part of the agreement, even though there's always the chance I, just like any major corporation, could be wrong.  I actually thought they would have welcomed my proposal.  But I guess they want me to play socialist while they play capitalist.

Forgot to mention that prior to the fix, whenever I used to click on the Autoswitching for graphics card in Energy Saver, it was always instant freeze and death.  Not any more.  Additionally, no more bursting with the mouse pointer or when scrolling down pages.  Just smooth cursor movement and scrolling.

<Email Edited By Host>

You want Apple to pay you for the "fix" to this issue? Good luck with that.

no.  i want them to compensate me to fix your broken 2011 mbp and everybody else's broken 2011 mbp experiencing these same symptoms.

BTW, have you ever considered trying to project how much Apple and mbp owners around the world have spent these past few years to "resolve" this huge issue?  Or how many Apple lovers may be lost forever?

Or that technicians even when they potentially misdiagnose a huge problem still receive compensation?

Have you ever considered that when you buy an expensive Microsoft or Apple product and you need their technical support, they expect you to compensate them to support the product you just paid them for?

You're not a socialist, are you? 

I actually believe that Apple hasn't spend a penny trying to resolve this issue. And I'm no socialist and never implied that I am. The only thing I ever said was that I doubt Apple will ever pay you for your "fix."

xslipper wrote:

 

no.  i want them to compensate me to fix your broken 2011 mbp and everybody else's broken 2011 mbp experiencing these same symptoms.

Don't hold your breath.

I have the exact same issue on my Macbook late 2011 15" with a dedicated AMD Radeon HD 6750M. Using gfxCardStatus doesn't resolve the issue even if I force it on integrated; I've had it crash on me with both dedicated and integrated. The only way to workaround the problem is to place the Macbook over a fan so that it keeps it cool. It's obv a major inconvenience.

I've already called Apple and they've opened a case for me, lets hope we can get an out-of-warranty resolution on this soon.

Just got off the phone with a Senior Service Rep from Ireland. They don't consider it a production issue and if your device is outside of the guarantee under local consumer law, you're out of luck. As I am a developer and OS-omnistic I can easily switch to Ubuntu and Dell XPS 13. By the way: Dell had a problem with that device ("coil whirl") and they're recalling and setting up the program now. What a difference! Unfortunately, you can only see how good a company is once things go wrong. Ciao and God Speed!

I got it back working, luckily I had 3 year warranty from the store it was bought. I hope it last.