27 iMac (2012) crashes and requires smc reset to turn on

Seven pages of posts on the same issue. Sometimes, it's a good time to look over the evidence and see where it's going.

Here's what I learned:

- The issue seems to happen to certain 2012 27” iMac i5 and i7 - but they're not the only machines affected.

- The problem is exacerbated by high-temperature rendering and use of Adobe software, like Lightroom, Dreamweaver, and Illustrator.

- The problem happens "all of a sudden" and doesn't get better by itself. The problem usually appears about two years after purchase.

- Replacement of some hardware, like the graphics or logic board doesn't necessarily solve the issue.

- Apple repairs have been inconsistent - including logic board, graphics board, and power supply replacement. No solutions seem to last.

- It's not a software problem.

- The problem varies in severity, from occasion to as many as ten times a day.

Certain fixes don't work consistently, or at all:

- replacing RAM

- installing CUDA, though that might solve things for a short time

- PRAM reset

- power supply replacement

- logic/graphics board replacement

Some have noticed:

- The problem frequently happens at high temperatures - like 80 degrees C

- TechTool suggests that the electrical output from the sensors may exceed an acceptable range.

- The longer an affected iMac is unplugged, the longer you can use your machine before the problem recurs.

Here's what I learned:

1. The problem is not isolated to this "vintage" of iMac. Some owners of 2013 iMacs are reporting the same issue. The problem starts to appear around two or so years after manufacture.

2. The problem is exacerbated by extended rendering, or extended use of graphics software like Lightroom, Dreamweaver, and Illustrator.

3. Once the problem appears, it doesn't seem to get better by itself. In other words, the damage has already happened (whatever that damage is).

4. For some (but not all affected machines), the pattern follows the classic appearance of a power supply failure.

There are lots of questions I'd like to ask affected users:

- How often do you use your fans? Do you use them aggressively?

- Do you use your Mac for long, intensive rendering work?

- Have any users tried fixing their machines with a technique that seems to fix with similarly-affected MacBooks - a BGA re-ball?

An Additional Update:

Since obtaining my system from repair (i.e., power supply and power cord replacement) over a week ago, I have not experienced an intermittent shutdown.

I suggest obtaining the shutdown code that is associated with an intermittent shutdown. In my situation, as soon as the Genius Bar representative had this code, he was able to make a diagnosis and propose a repair solution. Granted, I am still under Applecare and that may have been a factor in his little hesitation to repair the system. Prior to obtaining this code, the Genius Bar was unable to replicate the issue.

Finally, although I may or not be in the clear and that each situation has similarities and differences, I'm thus far satisfied with how Apple, specifically Matt (Genius Bar Representative) at the Northbrook, IL location, handled my system's issue.

And yet another one to be added to the list...

iMac 27-inch, Late 2012

3,4 GHz i7

32 GB RAM

3TB Fusion drive

NVIDIA GeForce GTX 675MX 1024 MB

The (sudden) shutdowns started the day after I updated the system to 10.10.6 and added RAM. The following was tried, to no avail:

-Replaced RAM with the original modules

-Clean install of the system (as a result of the 3TB hard drive replacement program)

-Joined the Apple Beta testing program to see if the newest El Capitan builds would help.

I can reproduce all of the behaviors that seem characteristic to the issue: sudden, random shutdowns; turning on only after pulling the plug for at least 5 minutes; the longer the plug is pulled, the longer the computer works; PRAM reset triggers the shutdown, etc.

Will take the computer to service this week (Amsterdam) and will reference this discussion. I would like to have the power supply replaced, as this is the most logical solution, according to my electronics-savvy friends.

this thread may be relevant to our interests My imac keeps restarting and freezes on grey screen with safe mode restart

Great list.

I've had the exactly same problem all while I was supposed to deliver a job. I'm a visual effects artist and was hitting the computer with renders and it was basically restarting every 15 mins.

my specs are like everybody else's

Mac (27-inch, Late 2012)

3.4 GHz Intel Core i7

NVIDIA GeForce GTX 680MX 2048 MB

After two weeks of multiple kernal panics, reboots, reinstalling the system, getting the apple store to replace the logic board, "downgrading" to Mavericks, replacing RAM and ultimately having to buy a new iMac to finish my work. I've since put the new computer aside and started over. I'm convinced it has to do with the GPU. I haven't installed the CUDA drivers and there fore my programs aren't putting any pressure on the GPU. It's running good so far. I'll post again in a week and a half after 18/hr a day rendering and use.

> I've had the exactly same problem all while I was supposed to deliver a job. I'm a visual effects artist and was hitting the computer with renders...

msolalski, your problem sounds a bit more like my GPU stress issue than the other power related issues in this thread.

For me, ALL my issues went away when did a clean install of the release version of El Capitan (El Cap betas had the same problems as Yosemite etc). I have run every stress test and software that caused the problem before and no longer get or can replicate the artifacts, crashes, panics, shutdowns etc. Occasionally when I run a 3d stress test under El Capitan there is a brief flash of something that looks like a video artifact as the test starts but it disappears and the test runs as normal. I can run the temp on the GPU all the way up to 90 with a game or stress test and have no problems at all.

I am not sure if there is still hardware issue and the new code/drivers are handling the issue better or if it was a driver/software issue all along. Right now I am trying to decide if I am going to sell the old mac to recoup the cost of the new one - I don't want to sell it if there is a hardware issue but with El Cap it is running perfectly fine. I am glad I waited a bit to get the logic board replaced at my cost...

Hi, I have had a very similar issue with my late-2012 27" iMac and have been following along here for a month or so.

27-inch, Late 2012

3.4 GHz Intel Core i7

16 GB 1600 MHz DDR3

NVIDIA GeForce GTX 680MX 2048 MB

Fusion Drive 1TB

Since I bought the system in early 2013 there were a few times my son had some freezing issues playing 3d games but never anything that wasn't fixed with a restart and I just wrote it off to driver problems or issues with the game.

Mid-summer this year I began to experience the system locking up with video artifacts filling the screen(small red and green squares mostly, sometimes others). It would happen during any system intensive task such as video editing or rendering, photoshop work, etc. 99% of the time I would have to reboot manually and even then most times the system would refuse to reboot, hanging with a few light graphic artifacts appearing during the grey screen with the progress bar. Resetting PRAM etc would not help and going into hardware diagnostics would not report any issues (even though the Hardware Diag. screens would occasionally have graphic artifacts or flashing colours themselves) The only thing that seemed to get things running again was leaving the the system off for a while and then rebooting. Over the course of a month I tried just about everything - different video drivers, installing the El Capitan beta, reinstalling a completely clean OS etc - without any any luck in resolving the issue. The odd time a panic log was generated it seemed to show the video drivers were the source of the problem.

I eventually found I was able to replicate the problem 100% percent of the time by running GpuTest's FurMark. As soon as the GPU Temp reached 70-80 C the system would hang with extensive video artifacts etc. It seemed like a classic case of a GPU or video memory failure due to a solder joint, very similar to the issues the led to the 2011 Macbook GPU replacement program. Everyone I consulted said that logic board replacement was my only option. Unfortunately it is not under warranty so the cost would be extremely high. I ended up pickup up a new iMac just so I could keep working, figuring eventually I would get the logic board replaced so I could sell or repurpose the system (and probably hoping Apple would announce a replacement program).

Now the interesting part - Wednesday I had some spare time so I replaced the El Capitan beta on that system with a clean instal of the full release version. Since then the system has been on full-time and I cannot replicate the issue at all, either through normal usage or running GpuTest/other system stress tests. FurMark has been running on that system almost constantly during each day with the GPU temp hitting 90-93 C. Occasionally when I stop and restart the test there is a very brief graphical glitch for a split second and then the test runs fine. Other than the updated OS the only thing I can think may be a factor is that the system had been turned off for 4-5 days before the El Cap install.

Any thoughts? If the issue does come back I will post an update.

Just jumping into this thread now. Having the exact same problems, showed up for the first time about a month ago after two years of usage. For me the shutdowns have occurred when working in FCPX. iMac is still under applecare but because the problem is so difficult to diagnose I'm struggling to find a resolution. Getting a little desperate here to figure this one out because my work is entirely dependent on this imac.

iMac 27 inch (Late 2012)

3.4 intel Core i7

32GB 1800 MHz DDr3

NVIDIA GeForce GTX 680MX 2048 MB

Fusion drive 2tb

Just an update - My problems with my iMac were not solved by El Capitan in the end. The problems are harder to replicate with system stress tests but I did get the hang/shutdown/artifacts a couple times on that system while my son was playing a lego game over the weekend. I was hoping the issue was solved as I wanted to put that iMac up for sale to recoup the cost of my new one. Oh well...

Hi, I have had a very similar issue with my late-2012 27" iMac and have been following along here for a month or so.

27-inch, Late 2012

3.4 GHz Intel Core i7

16 GB 1600 MHz DDR3

NVIDIA GeForce GTX 680MX 2048 MB

Fusion Drive 1TB

...

Hi guys so I've been having this same issue before and was in discussion in the earlier posts. Just now my iMac died... I was giving a one last try after 1 week not using it and it just crashed on me. Did a SMC reset and still no respond. It is officially dead. For a $3000 dollars computer that doesn't even work for more then a year sounds like a manufacture problem to me. I have a 10+ years old laptop and it still works perfectly fine. To be honest, I'd really rather having a slow computer than one that crashes every 30 minutes and dies on you after a year of usage..

I felt ripped off and justice should be done. Is there anyone planning to file a law suit against Apple? If so please contact me, I'd love to join.

The computer has been back from service for three days now, where they replaced the motherboard.

The technician who performed the job summarized the issue as a "Power unit/motherboard communication failure after a certain temperature was reached". Too bad he wasn't more specific but it could be related to some motherboard's component's latent (ESD?) damage triggered by expansion due to the temperature rise.

My computer's motherboard was replaced free of charge ⚠ even though the computer was purchased in March 2013 and I have no Applecare. This was possible due to the Dutch Consumer Law that basically says, very subjectively, that "all consumers are entitled to a proper product".

I hope you guys are able to solve this in any satisfactory way. I'm happily signing-off this thread...

Update:

I shouldn't have ended my previous post so triumphantly... the computer has randomly shutdown twice in four days after the logicboard replacement. This looks less frequent than before the LB replacement but the feeling is the same: can't trust the computer.

Will take the computer back in this week, probably for a new SMC. Hopes that it will solve the issue are not high, though.

Apple, I believe the ball is on your court, now.

Yet another customer here who has been ripped off and now has an expensive piece of junk that is basically unusable and no solution whatsoever from Apple that will actually fix the problem. Come on Apple, this thread and many others like it surely points to the fact that the late 2012 27" imac is a complete lemon and needs a recall. I've been a loyal and happy Apple customer up until now. Incredible, I was actually able to type this post before my imac shut itself down...

Worst thing is that I can't even sell the **** thing and buy a new (hopefully more reliable one) as who would buy a 2 year old imac that is in "excellent condition" apart from the fact that you can't guarantee that it will actually operate for longer than 30 minutes at a time...

iMac 27 inch (Late 2012)

3.2 intel Core i5

16GB 1600 MHz DDR3

NVIDIA GeForce GTX 680MX 2048 MB

A solution?

BGA reball. Look it up.

I looked it up and saw no reference to it being used as a solution to the problems being discussed in this thread. Are you saying you have a faulty 27" imac and have used this as a solution? Please elaborate. As much as I currently hate my Mac, I'm reluctant to take a soldering iron to it...

Perhaps my suggestion was a but enigmatic.

Given my reading and experience with this issue, coming up with a "fix" is a real problem for some users. Affected users have tried changing out logic and graphics boards, etc.... but the one fix that seems to work every time is the BGA re-ball. (And not the kind that comes in a kit.)

This is not the same thing as a solder reflow, though the basic principle is similar.