Logic Board Failure due to Nvidia Graphics processor GT330M

"Sam

@leafixrepair


Rep: 1.9k




7


6


2"

"Kernel Panics are often corrupted memory blocks caused by a misbehaving process. While it's possible the cause could be bad hardware (RAM or the GPU) I've seen more software causes than hardware failures.

Did you upgrade your OS or add/upgrade any applications just before you started seeing the kernel panics? There have been reports of some video driver problems during the Lion upgrade process that cause odd kernel panics.

I would first try isolating out the software causes. The easiest is by using an external HD (Firewire would be better than USB here) which has a fresh copy of Mac OS-X 10.6.8 on it (not Lion 10.7.x if that is what you had been using). Then select it as your boot disk (you can still use the internal HD's apps). Did the problem go away? If that did make a difference you'll need to make a full backup of your HD wipe it down completely then using an external HD to boot from install a fresh copy of OS on the internal and use the migration tool to copy back your files.

One note here: if you did upgrade to Lion you want to boot under a straight Lion OS external HD not one had was upgraded or copied from an upgraded Leopard or Snow Leopard OS (to make sure you don't copy over any of the older drivers). To prep and install the fresh copy of OS.

If the panics are still happening after booting up under the external HD you could try exchanging out the RAM with a fresh set of SIMM's. I wouldn't spend the effort moving your current SIMMS from slot to slot or removing one, unless you upgraded the RAM after you bought your system (different manufactures). Otherwise I suspect one of your Apps has an issue, remember what apps you use and the order you use them after a fresh reboot to see what apps could be suspect then focus on making sure you upgrade these to the newest versions (don't forget OS add-ons and virus checkers).

As far as the Nvidia issues of the older MB Pro's ('08) Nvidia had a bad run of chips do to a design flaw. Depending on where your chip was located on the wafer the thickness of the layers are very slightly different. The inner core are thicker and the outer edge are thinner (think how centrifugal force works). The current theory is the dies (chips) closer to the edge just couldn't hold up as well under stress as the others. Normally these outer edge chips get tested more fully and depending on what happens they are tossed out. Some how the process failed so some got released. As far as I know only the 8600M GT was effected. As your system has the 330M GT it does not relate as being the same problem.

Check this out about the older Nvidia 8600 GT chip: http://www.macrumors.com/2008/10/10/some...

Bottom line here is your issue might be a new issue with the 330M GT chip which so far has not been acknowledged by Apple or Nvidia. So far no other system manufacture has had issues with this chip either as far as I can find.

References:

http://www.notebookcheck.net/NVIDIA-GeFo...

http://www.notebookcheck.net/NVIDIA-GeFo..."

"All of you are wrong, the problem and resolution is here:

https://www.youtube.com/watch?v=DzcgT_fi..."

"You're going to love this...

It is very probable that issues with panicky GT 330M's isn't actually a problem with the GPU as many have thought.

According to this amazing thread on MacRumors:

""Apple made the same mistake they made with the 820-2610 and 820-2330 where they ran out of space on the motherboard for C7771 so they made it a tantalum one. This is not suitable for a rail that is always on, as C7771 is for an S5 rail - so it dies. Voltage fluctuates, and eventually the machine stops turning on. On the 820-2850, that same 330uf tantalum capacitor re-appears on the framebuffer power supply for the GPU and the VRAM.''

One possible fix is to replace the dodgy cap with a regular type, however in the mentioned thread you'll find a soft workaround, which involves patching the Info.plist of AppleGraphicsPowerManagement.kext.

And the best bit....

I got my GPU Back!!"

"I had the same desease for a long time.

The bug is hardware : it's due to a bad type choice on a decoupling capacitor used on GPU motherboard part

But good news, you don't have to replace the motherboard

Julian Poidevin sent me this:

Hi, I created a software to automatically correct this issue, you can download it on github, it's free :

https://github.com/julian-poidevin/MBPMi...

Install it. Password requested is your login session password because that software must change SIP (System Integrity Protection) to force GPU to take medium speed always. For more info read FAQ in poidevin site.

Now my MBP 15 mid2010 is working fine

If you have installed GFxCardStatus let it in dynamic mode.

Any time you update MacOs you must do a reinstallation of the poidevin fix again.

That fix was developed based on fabioroberto investigation

https://forums.macrumors.com/threads/gpu...

see post #20"