Q: 2011 MacBook Pro and Discrete Graphics Card

 

That should mean that after replacing the logic board, we would get another two years of usage without any problems. But the truth is that I had my logic board replaced three times already (after paying an 500euro replacement repair, which is about $700 us) and none of the times the computer lasted more than a day without giving me problems. All three times I updated the software following the normal procedure (App Store) which implied upgrading the firmware. All three times the computer was working before the upgrade and broke after the upgrade. I do think this is the real cause of the problem. Otherwise I would expect to get another two years of usage until the computer would break because of the design flaws mentioned earlier.

The problem is: There's a high chance that you don't get a new logic board. When they are replaced they'll be likely be sent back to Apple (out of curiosity and to prevent it from being refurbished I wanted to keep my defect logic board but then the price quote took a hike from 590€ to 1092€ so I declined) inspected, if (deemed necessary) fixed and then they'll likely run a stress test and ship it back to stores as replacement part. The problem here is that they might not exhibit the problem when not installed in a case due do different strain. So they might be reused although they're defect replacements from other customers. It's not the first time I'm seeing electro-mechanical problems with ATI chips, in fact I "fixed" a friends iBook by simply putting a piece of cardboard in applying pressure on the discrete GPU. Although this problem seems to be more widespread with NVidia because there've been far more recalls I had this problem with exactly every Apple I bought all relying on ATI chips -- in fact I bought my last Mac despite ATI graphics only because I though they had a grip on the problems now thanks to NVidia-gate. Seems I was wrong.

If that is the case then they lied to my face by saying that all the times I got a brand new replacement.

degger wrote:

 

 

That should mean that after replacing the logic board, we would get another two years of usage without any problems. But the truth is that I had my logic board replaced three times already (after paying an 500euro replacement repair, which is about $700 us) and none of the times the computer lasted more than a day without giving me problems. All three times I updated the software following the normal procedure (App Store) which implied upgrading the firmware. All three times the computer was working before the upgrade and broke after the upgrade. I do think this is the real cause of the problem. Otherwise I would expect to get another two years of usage until the computer would break because of the design flaws mentioned earlier.

The problem is: There's a high chance that you don't get a new logic board. When they are replaced they'll be likely be sent back to Apple (out of curiosity and to prevent it from being refurbished I wanted to keep my defect logic board but then the price quote took a hike from 590€ to 1092€ so I declined) inspected, if (deemed necessary) fixed and then they'll likely run a stress test and ship it back to stores as replacement part. The problem here is that they might not exhibit the problem when not installed in a case due do different strain. So they might be reused although they're defect replacements from other customers. It's not the first time I'm seeing electro-mechanical problems with ATI chips, in fact I "fixed" a friends iBook by simply putting a piece of cardboard in applying pressure on the discrete GPU. Although this problem seems to be more widespread with NVidia because there've been far more recalls I had this problem with exactly every Apple I bought all relying on ATI chips -- in fact I bought my last Mac despite ATI graphics only because I though they had a grip on the problems now thanks to NVidia-gate. Seems I was wrong.

The difficulty in assigning blame to firmware updates (or any other time-related cause) is that users have been documenting these issues since 2011 and there doesn't seem to be any sign of obvious causality.  People have pointed at every OS update and firmware update as a potential culprit, but there hasn't been one that seemed to be followed with a spike in reported problems.  There have been periods of heavier and slower activity in these threads since they started.  The results from the survey I created in July are heavily skewed by the fact that most of the people who have responded to it are newly-affected users from this summer and fall - most of the users affected earlier didn't stick around long enough to see the survey and respond.  There's very high turnover in these threads, which is unfortunate for information gathering and knowledge retention.  Even though my problems are solved at the moment by logic board #4, I am still reading these threads because I don't want to lose sight of the bigger picture. 

Thank you for staying with this, saramwrap.  I think this issue would be a lot harder to follow without your efforts to bring a sense of continuity to the discussion.  It was nice to see that you were mentioned by name in the recent articles. 

I run a repair company in the UK and we have recently been asked more commonly for repairs to the 2011 ATI Macbook Pro's.

Before I start this is not an advertising campaign but just my input to clear up some things and offer some advise.

We specialise in replacing GPU's. We do GPU repairs everyday and have done since 2008.

At first back in 2008 we had many of the Pre-Unibody 2007/08 units coming in for repair with the 8600M GT. This was of course later covered by Apples Extended Logic Board replacement programme.

We first started reflowing - Result works for 3 to 6 months generally, if lucky. Result - Units would come back.

We then progressed to leaded reballing. Results were better but machines were still coming back. General repairs would last 3-12 months but usually a little longer than a reflow.

In 2009 Nvidia released a revised GPU. This was seen in the REV.2 board. We sourced and started fitting these with leaded solder for extra reliablilty and safer soldering temps.

Result - The units were not coming back and to date we have only seen a very small handfull of machines come back. The majority have been working for years.

The problem with the 2007/08 Nvidia GPU's was the unsuitable materials used to produce it causing the chip to fail internally. Once it failed heating would restore the connection for a few moths but it would always fail.

My intial thoughts are that ATI Chipsets in the 2011 models may have problems internally. So if you are going to have a repair make sure the company fits a brand new chipset reballed in lead solder - WITH PROOF!

We always fit brand new chipsets for all GPU's. Returns are not good for business and we do not want to see our customers again as they will be unhappy ones!

We have repaired 4 of these units so far. Brand new chipsets fitted and the longest was around 6 months ago - A good initial sign.

If anyone needs any more advise feel free to ask me. We advertise on ebay but only for the A1226 A1260 ... etc models so will be putting up a repair service for these soon when we have done more. It would be nice to repair some users machines here so we can get some feedback on reliability and some more of these 2011 repairs under our belt.

To note we change GPU's everyday and rarely make mistakes so our repair rate is 98%+ for non previously butchered units... Our 2011 is 100% apart from 1 that had been damaged with a heatgun.

degger wrote:

 

 

...It's not the first time I'm seeing electro-mechanical problems with ATI chips, in fact I "fixed" a friends iBook by simply putting a piece of cardboard in applying pressure on the discrete GPU. Although this problem seems to be more widespread with NVidia because there've been far more recalls I had this problem with exactly every Apple I bought all relying on ATI chips -- in fact I bought my last Mac despite ATI graphics only because I though they had a grip on the problems now thanks to NVidia-gate. Seems I was wrong.

Heh.  I employed the same "fix" on an iBook thinking that I would eventually get around to fixing it correctly, but it never gave me another problem  If only everything was so easy!

I think it's interesting that, unlike the Nvidia incident, we're not seeing the failure of these AMD chips in other brands.  I wonder why?  I just spent about 30 minutes on Google trying to find a single example of a 6000 or 7000 series chip failing in something other than a MPB and came up empty.  There have been some driver issues, but nothing looks outside of the norm to my eye.

This makes me even more curious.  Does anyone know whether or not these chips come pre-balled when they're ordered in large quantities by companies like Apple?  That's they way they would come if you or I ordered one from a parts vendor.  Or is it that optional in manufacturing?  Can a large player like Apple request that a certain type of solder balls be used?  Or none?  Or is it uniform across the board?  I just wonder if we're talking about the same solder in all of these other brands too.  If so, I suppose that would remove some suspicion from the solder (As a noun, but not as a verb--There are still a lot of variables in assembly that can effect the thermal fatigue properties and longevity of lead-free solder joints). 

This makes me even more curious.  Does anyone know whether or not these chips come pre-balled when they're ordered in large quantities by companies like Apple?

If you order fully packaged chips they will always come with proper their designated type of connectivity. In this case the chips are usually BGA (Ball grid array) so they'll come with solder balls. There sometimes is the possibility to order chips in forms not ready for direct use in manufacturing but those will typically be used for die-stacking.

Can a large player like Apple request that a certain type of solder balls be used?

Typically there're several specifications of chips available for different usecases, like military spec or unfriendly environment spec which will typically have a different solder material to match the requirements. In many parts of the world lead-free manufacturing is required by law so Apple needs to adhere to that. There're not too many suitable lead-free alloys available, especially not cheap ones so the choice is usually limited.

I just wonder if we're talking about the same solder in all of these other brands too.  If so, I suppose that would remove some suspicion from the solder (As a noun, but not as a verb--There are still a lot of variables in assembly that can effect the thermal fatigue properties and longevity of lead-free solder joints). 

Indeed, but as someone else noticed there're no reports about non-Apple products failing with these AMD chips so this seems to be a manufacturing process or design problem.

degger wrote:

 

Indeed, but as someone else noticed there're no reports about non-Apple products failing with these AMD chips so this seems to be a manufacturing process or design problem.

Aye.  And there are a LOT of variables there.  There have been hundreds of academic studies published over this past decade just on the subject of how the cooling rate shapes the strenths and weaknesses of the bond formed. 

I think we'll get the kinks worked out with lead-free solder pretty soon.  The electronics industry should probably have been allowed a longer transition period.

degger wrote:

 

No, you want to inquire about a GPU replacement, not a reball and certainly not a reflow. If done right a reball is a high risk high cost measure that only makes sense if you really need to preserve a particular chip at any price.

I don't feel as strongly about that as you do, so I'll leave that for readers to research further (anyone considering these alternative routes really should do some research and not take my posts here as the final word about anything). 

However, for board-level repair work, I do favor the use of traditional lead based solder balls (Sn₆₃Pb₃₇).  Even if you decide on a GPU replacement, I'd still consider reballing.  I would at least ask about it.  Lead based solder has a lower melting point (and better wetting characteristics) than that which was used in manufacturing, so less heat for the second part of this procedure.  Also, it has a good reputation spanning decades for its ability to withstand thermal cycling (that's not just turning your computer on and off, it's all the dynamic switching between GPUs and even just changes in your minute to minute computing activity).  It's easy for a small operation to work with and you get a very nice bond at much lower temperatures. 

Also, too, NO WHISKERS! 

Do some research and then find a good facility and inquire about these options there.  Try to make it a point to speak with a actual nerd (not a customer service person and certainly not one of those creepy managment types). 

Again, though, it might be wise to hold off a while longer.  As mentioned many times in this thread, any of these alternative options could void future eleigibility for a repair program if one is created.

However, for board-level repair work, I do favor the use of traditional lead based solder balls (Sn₆₃Pb₃₇).  Even if you decide on a GPU replacement, I'd still consider reballing.

I find it rather interesting that you ask whether chips always come with balls to big vendors but then go on to recommend a reballing on a brand new GPU...

The point of soldering a new GPU is exactly to avoid the reballing which is a tedious and risky (and/or expensive) operation. If you're going to reball you might as well use the old GPU chip because the whole replacement procedure is exactly the same as doing so while reballing a new GPU.

Companies nowadays have very good experiences with lead-free solder, only Apple seems to have messed up with this particular chip; potentially because they reflowed using unsuitable parameters. When placing a new GPU and properly reflowing it that should work much longer, potentially longer than a shoddy reball job with lead based solder. The problem here is finding a real "pro"-shop to do it.

It's easy for a small operation to work with

Absolutely, if you're talking about the lead based solder. Reballing however is absolutely not.

Again, though, it might be wise to hold off a while longer.  As mentioned many times in this thread, any of these alternative options could void future eleigibility for a repair program if one is created.

True, or in the likely case you need the machine until Apple decides to do so (worldwide I might add, not the US-only law suit mitigation kind), get the board replaced by Apple or an ASP.

Using gfxCardstatus as temporary fix but unfortunately I had to shut down and now I am having no luck booting. Used to work within an hour but not anymore. Even in safe mode. I've been attached to this board for quite some time and have gone through a lot of the tricks to get it to boot (cool down --> single user mode --> exit, let it go on grey screen for a bit then reset, plugged in, not plugged in, pram reset, spiritual dances, etc). I'm now at the point where I want to move the ATI files to the desktop but I can't for some reason. Following some instructions on this and other forums, after booting into single user mode I,

mount -uw /   --> make files and folders read writeable

cd /System/Library/Extensions  -->  go to the directory

mv ATI* ~/Desktop  --> move ATI files to desktop

here I get,

Usage: [-f  |  -i  |  -n] [-v] source directory

           [-f  |  -i  |  -n] [-v] source ... directory

The files don't move. I'm not a unix guru but I'm pretty sure this has something to do with permissions and/or what I am typing. I have tried to move non system files with some success: I moved a .pdf file to the "Desktop" only when I go to the desktop directory I do not see the file there, implying its moving to another "Desktop" folder which may explain some of the problem. 

If anyone out there more savvy than I can let me know how I can move these files successfully or inform me of the voodoo that they do to simply get into the operating system so I can switch to integrated only, it would be greatly appreciated.

Thanks!

I have started using gfxCardstatus and forced it to use Intel Graphic card and this is was the first time ever I got this issue in 2.5 yrs of this laptop and I think they should really bring out some fix, many people are being effected by this then why isn't apple looking into the matter and this form has grown so much still they havent looked into the matter.
Apple isn't the same after Steve Jobs Death it has been making so much problems with the software and hardware that this may lead them to a really bad shape in future. This company use to look to make things better, not ruin for people.
Hope they look into the matter ASAP now and bring the fix soon and maybe bringing out this form to other big sites and they posting this will make Apple to look into the matter. Hope for the best.

I have same problem.

I have Macbook Pro 15" early 2011 2.0 GHz

I have same problem!!

MBP 15" 2011 Late

I am so sad

This is horrendous. How has Apple STILL not answered this at all? This shows absolutely no accountability at all. I will be making sure that every person in the industry that I know is notified of this. If Apple thinks that the "nerds" buying laptops to write software for their platform aren't important, they have another thing coming. There's absolutely no way that any software engineer with financial sense will spend $3000 on a computer with hardware that can't last two years. I'm embarrassed to see Apple rated so highly in customer service with the awful experiences I've had just in the past month.