Q: 2011 MacBook Pro and Discrete Graphics Card

Okay, so what's going on here?

This is how it looks from where I'm sitting:

The central hub of everything being discussed here is the connection of the GPU to the board, right?  Nothing more.

The  fact that some folks have "fixed" the problem, even for a week, with a  reflow using an oven or with professional tools that focus heat  specifically on the GPU seems to be pretty clear evidence that we're dealing with  a solder problem.

Everything else is secondary.   How many hundreds of pins are there on a 6400 series GPU? They're all  connected using the same solder on the same board.  There is no reason  to assume that they are always going to fail in exactly the same way.   There will be some variation in the symptoms and the workarounds to  those symptoms.

It may or may not also be the case  that a particular OS/software/firmware/hardware combination will  highlight some variations of this manufacturing defect more than  others.  Enough people have mentioned it, that it probably shouldn't be  dismissed out of hand.  Given that some variation is to be expected,  we probably shouldn't find it surprising anyway.  But it doesn't matter.   This is all secondary and revolves around the same thing.  It's a  manufacturing defect. 

The cause:  Poor quality solder subjected to high tempatures over  multiple thermal cycles have caused fractures in the soldered  connections (Once those fractures form, problems can then occur at any  temperature).

The solution:  Yet to be determined by Apple (known for doing the right thing in the past, but is starting to look a little sketchy on this one).

Does that pretty much sum up the current state of affairs?  Or did I miss something when reading through this epic thread?  Did I get anything wrong?

Thank you.  I couldn't agree more with this assessment.  I'm a month into a professional reflow and have been 100% successful so far.  The mere fact that reestablishing solid connectivity through GPU solder points makes all the problems disappear is compelling evidence in my opinion. My machine will likely not last very long with the original solder, but the lesson remains the same.

I think Apple needs to get a metric ton of new logic boards, have them all resoldered to military specs, and then start swapping boards at no charge.  That, or they need to buy back these machines with healthy credits toward replacements.

A month?  I hope that continues.  Are you using the discreet or trying to stay with the integrated?  I've wondered about this from others.  I wonder how a reflow would hold up if the discreet were no longer used.  You're talking reflow, not reball, right? 

I hope it continues too.  I'm not restricting use at all.  I need an external thunderbolt display the majority of the time, constantly run Parallels with Windows 8, am on Mavericks, and generally push the machine fairly hard on a daily basis.

Yes, my "repair" was a $150 reflow with a 60 day warranty.  They also cleaned out ventilation and applied new thermal paste. 

I put the over-under at 90 days, although I wonder every day if this will be the day.  Will report when it does, of course.

I'd think that no longer using discreet would put less stress on the gpu solder.  The heat cycles cause the expansion and contraction that stresses the solder points.

Wise to take it to a professional.  With the right equipment, you can safely target more heat to a specific area than you ever could to the entire board in an oven.  It gives you the best chance of some longevity.  Best of luck to you.  As you know, even a professional reflow is not usually a permanent fix.

I had my screen replaced (for free) with my third logic board replacement. My issues start the moment I install the latest SMC firmware upgrade (1.8). Now my computer is sitting at the Apple Store for a fourth time and I am already considering buying a new laptop.

About SMC Firmware update, I found that on my latest replaced logic board, the SMC firmware version is 1.69f3 (MacBook Pro 8,2, 15" early 2011) while on Apple website (http://support.apple.com/kb/ht1237) the latest version mentionned for this model is 1.69f4.

But no SMC update 1.69f4 is offered when I check for updates, and if I click the link on Apple website it redirect me to a page mentionning a 1.8 version for mid 2012 MacBook Pro (so not for early 2011).

So it seems Apple removed 1.69f4 from download and from their software update server. If this is the case, this probably mean that there is indeed a known issue with that version.

For now in my case, so far so good on my replaced logic board. I hope it'll last.

No failure for me. Still on Snow leopard 10.6.7 but my battery died and my computer slow sometimes.So I reboot and after a while it restart to slow. I'm a graphic designer only using Illustrator, Photoshop, a web browser and a font manager on my laptop (only!)

Nom du modèle :          MacBook Pro

  Identifiant du modèle :          MacBookPro8,2

  Nom du processeur :          Intel Core i7

  Vitesse du processeur :          2,3 GHz

  Nombre de processeurs :          1

  Nombre total de cœurs :          4

  Cache de niveau 2 (par cœur) :          256 Ko

  Cache de niveau 3 :          8 Mo

  Mémoire :          16 Go

  Version de la ROM de démarrage :          MBP81.0047.B0E

  Version SMC (système) :          1.69f1

  Numéro de série (système) :          xxxxxxxxxxxxx

  UUID du matériel :          B23ECF48-BCC9-5F50-ABEC-5D22909AA728

  Capteur de mouvement brusque :

  État :          Activé

AMD Radeon HD 6750M :

  Jeu de composants :          AMD Radeon HD 6750M

  Type :          Processeur graphique (GPU)

  Bus :          PCIe

  Longueur de la voie PCIe :          x8

  VRAM (totale) :          1024 Mo

  Fournisseur :          ATI (0x1002)

  Identifiant du périphérique :          0x6741

  Identifiant de révision :          0x0000

  Révision de la ROM :          113-C0170L-521

  Version gMux :          1.9.23

  Version du gestionnaire EFI :          01.00.521

  Moniteurs :

LCD couleur :

  Résolution :          1680 x 1050

  Profondeur de pixels :          Couleurs 32 bits (ARGB8888)

  Moniteur principal :          Oui

  Miroir :          Désactivé

  Connecté :          Oui

  Intégré :          Oui

Connecteur pour le moniteur :

  État :          Aucun moniteur branché

islabonita- You are using a much older firmware! Mine is newer, but not the last one, and I have no problem too:

Hardware Overview:

  Model Name:    MacBook Pro

  Model Identifier:    MacBookPro8,2

  Processor Name:    Intel Core i7

  Processor Speed:    2,2 GHz

  Number of Processors:    1

  Total Number of Cores:    4

  L2 Cache (per Core):    256 KB

  L3 Cache:    6 MB

  Memory:    8 GB

  Boot ROM Version:    MBP81.0047.B27

  SMC Version (system):    1.69f3

MacBook Pro (15-inch Early 2011), Mac OS X (10.6.8)

If you try software update do you see a newer version for your computer?

iFrodo- I can't update too. It really seems that Apple have pulled this firmware from the servers. Maybe they, at least, are investigating this issue.

Until now, (that I know) everybody that has this older firmware still free from these glitches.

I have a more recent version of the firmware and I have the problem. Model identifier is also more recent. I see also that my version of GPU Rom is more recent than the one posted by islabonita

Model Name:          MacBook Pro

  Model Identifier:          MacBookPro8,3

  Processor Name:          Intel Core i7

  Processor Speed:          2,2 GHz

  Number of Processors:          1

  Total Number of Cores:          4

  L2 Cache (per Core):          256 KB

  L3 Cache:          6 MB

  Memory:          8 GB

  Boot ROM Version:          MBP81.0047.B27

  SMC Version (system):          1.70f6

AMD Radeon HD 6750M:

  Chipset Model:          AMD Radeon HD 6750M

  Type:          GPU

  Bus:          PCIe

  PCIe Lane Width:          x8

  VRAM (Total):          1024 MB

  Vendor:          ATI (0x1002)

  Device ID:          0x6741

  Revision ID:          0x0000

  ROM Revision:          113-C0170L-573

  gMux Version:          1.9.24

  EFI Driver Version:          01.00.573

iFrodo wrote:

 

About SMC Firmware update, I found that on my latest replaced logic board, the SMC firmware version is 1.69f3 (MacBook Pro 8,2, 15" early 2011) while on Apple website (http://support.apple.com/kb/ht1237) the latest version mentionned for this model is 1.69f4.

 

But no SMC update 1.69f4 is offered when I check for updates, and if I click the link on Apple website it redirect me to a page mentionning a 1.8 version for mid 2012 MacBook Pro (so not for early 2011).

 

So it seems Apple removed 1.69f4 from download and from their software update server. If this is the case, this probably mean that there is indeed a known issue with that version.

 

For now in my case, so far so good on my replaced logic board. I hope it'll last.

Sadly WebArchive isn't very complete but it looks like the firmware was pulled some time after May 2nd this year. http://web.archive.org/web/20130502234657/http://support.apple.com/kb/DL1633

While I'd love to think it's a firmware issue and rolling it back/getting a board that hasn't had it applied I'm still of the opinion it's a hardware manufacturing fault.

A forum member on here (Saramwrap) ran a survey, here's the analytics from it

https://docs.google.com/forms/d/1PuJ2QS2Ri7fw9HVaNHLQ4OYZCb99zonc96HmhTWEAoY/vie wanalytics

The 1.7 firmware was made available early Feb 2013 however the analytics show that issues were being reported before. My guess is that as the devices became older they started failing and not a firmware issue, or the firmware exacerbated the problem at best.

Wojzo- I totally agree with you. This is not just a firmware issue.  I am following this issue since last year when started to show for some people, before the last firmware was issued.

Since the beginning of the Saramwrap survey I am following the reports trying to find a pattern, but looks like there are a lot of variables in this equation.

This old thread has a lot of information since this 2011 mbp was launched:

https://discussions.apple.com/thread/2768351?start=0&tstart=0

I am suggesting that the last firmware just catalyzed the problem. Reducing the lifespan of the components with some sort of voltage rise causing overheating that could compromise the chip and the Vram. Overclockers already attested that the 6750m chip can run stable with less voltage applied:

http://forums.macrumors.com/showthread.php?t=1142028

I don't know why they would rise a voltage on this chip, but this is a strong factor to reduce a component lifespan.

By the way, on the 'software update' the 1.7 firmware was showing until July, the last time I checked this thing.

About SMC Firmware update, I found that on my latest replaced logic board, the SMC firmware version is 1.69f3 (MacBook Pro 8,2, 15" early 2011) while on Apple website (http://support.apple.com/kb/ht1237) the latest version mentionned for this model is 1.69f4.

The replacement board for my 8,3 came with

SMC Version (system):          1.70f3

@Degger: Do you have a 15 or 17" early '11 MBP?