MacBook Pro late 2011 had GPU problems

gpuhardwaremacbook propower-management

I have a MacBook Pro late 2011 and yesterday my external screen suddenly showed glitches on the screen; much like the screenshots you see from other 2011 MacBook Pro's that have a failing GPU.

As I work with the lid closed, I opened the lid and closed it again and the problem would disappear for a few minutes. I then disconnected my external screen and tried my MacBook pro's build in screen. After working with that for an hour it also started to show glitches. These did not go away after opening and closing the lid.

Becoming more and more annoyed I decided to reset the PRAM, and low and behold, it didn't work. So I went on with resetting the SMC. This is when the real problems started. My MacBook pro now booted up into a grey screen after loading for a bit.

Using Single-User mode I tried repairing the SSD with fsck -fy, this didn't help. Slowly I had to come to the conclusion that it most likely was a dedicated GPU hardware failure.

Then, suddenly my MacBook pro decided that it would want to boot up in Safe Mode, however, after a few minutes vertical and horizontal lines showed up again.

So, I decided to turn off the dedicated GPU by moving the kernel extensions into another folder. This worked but made everything very slow. After putting everything back, I decided to follow the following instructions:

http://www.asyncro.com/2014/03/24/macbook-pro-discreate-graphics-card-issue-fix-updated/

Having followed every instruction, including letting my MacBook overheat, my notebook was still slow.

Now, here is the thing, after putting all kernel extensions back into place and restoring everything back to normal, the notebook suddenly started normally at full speed using the dedicated graphics card.

Until now I have not experienced any problems (few hours later). At first I thought that it might be because of the gfxCardStatus software, so I deleted that in order to replicate the problem. However, my MacBook is still running fine now and I seem not to be able to replicate the problem.

My Question

  1. Can it be that all my problems were caused by a badly behaving SMC and that there is nothing wrong with the dedicated GPU?

Best Answer

No, you most likely have the very same common GPU issue. Seeing the glitches multiple times is a sure tell sign.

Beneath all that drama caused by resetting and tweaking things, temperature variations on the GPU contacts were what caused the glitches and the inconsistent behaviors. Your laptop is probably in the initial phase of GPU(soldering) failure. Sooner or later the problem worsens and more severe graphic glitches will appear or the computer simply becomes unbootable.

You can try the simple steps below to stress-test the GPU, but by doing so you could be accelerating the problem and end up in an unusable state sooner. You can decide if it's in your interest to forcefully trigger the problem. Regardless, these steps are not beyond any normal operating parameters, so if your laptop is immune from the GPU problem you are not doing anything harmful by performing these steps, warranty-wise especially.

  1. Get a temperature & fan speed monitoring app like iStat Menus that can show you fan speeds, CPU and GPU temperatures in real time.

  2. Run a GPU benchmark app to stress the GPU. Valley or Heaven should work, for free. Keep running/looping so the GPU temperature stays around 70'C or higher, for as long as possible (hours).

  3. You should pause the benchmark app periodically for the temperature to drop down (back to 45~50'C range). Or just turn it off at nights. Repeat the heating/cooling process until the symptoms appear.

This test might end up proving nothing, and yet it doesn't mean the GPU is trouble-free. It takes time and many temperature cycles for the solder joint problem to become apparent.