Q: 2011 MacBook Pro and Discrete Graphics Card

Please understand that I am looking at this as a way of gathering information, it may result in a fix for you but equally, it may not. But until enough people actually provide information it will be impossible to see the pattern (if there is one)

Let's use some logic. AFAIK no Mac with IGPU only has shown such issues. On the other hand rebelling/logic board swap has fixed the problem for lots of people here, albeit for a limited amount of time in some cases.

Of course non stock ram and optibay solutions are not supported by apple and of course a second HD adds heat. Is that heat enough to completely screw the thermal balance of the machine? Even a 7200 RPM model shouldn't add much heat, not even close to the heat rise from winter to summer (which the computer should be made to withstand). Also, previous Macs with DGPU have had problems. I used to own a 2007 17" non unibody (Nvidia 8600) which had the exact same issue and required a logic board swap (fortunately the replacement boards had a new GPU revision which eliminated the issue).

Finally a friendly apple genius told me some moths ago that every Macbook Pro with DGPU is subject to an extreme thermal stress which can cause the system to go haywire. This is why when I'll buy a new MBPro I'll also get Apple Care and will sell it as soon as the warranty coverage ends.

Hi GPU power coupled with poor ventilation will always produce issues.

FYI, over two weeks ago, I replaced the thermal paste on my GPU and CPU.  Since then, my MBP 2011 has been running every day, heavy use and sleep mode, with not one graphical glitch.  I think the key in my situation was replacing the paste soon after first noticing the freezing.  I think that prevented heat damage to the chips.

post.user_wrote.label:

 

FWIW, my 2011 15" MBP began experiencing this issue last week.  More and more, the system would freeze, and my screen would turn blue or red with vertical lines.  I found this thread and started reading up on it.  Long story short, I fixed the issue as others have suggested, by replacing the thermal paste.  My AppleCare ran out so I decided to take a chance.  I wouldn't do this unless you're ready to void your warranty, and\or risk killing your hardware.

 

1. I took apart the MBP (following steps from the ifixit teardown guide, though it skips a few steps)
2. Removed the heatsink from the CPU and GPU
3. Scraped\cleaned off the thick paste with rubbing alcohol and lint-free non-abrasive wipes
4. Re-applied a thin layer of Arctic Silver Alumina ceramic thermal paste (using this method)
5. Re-attached the heatsink
6. Put the MBP back together
7. The entire process took about an hour
8. Allow 36 hour break-in period of thermal paste - NO FREEZES AFTER THIS

 

What's interesting, is ifixit foreshadows heat issues in their teardown guide:

 

Holy thermal paste! Time will tell if the gobs of thermal paste applied to the CPU and GPU will cause overheating issues down the road.

 

They say thermal paste should be replaced every 2-3 years, so I'm convinced if you identify the issue soon enough, you can fix it before the chips are damaged.  Only time will tell if my machine is truly fixed, however after running graphics intensive programs and games for hours, I have not gotten one crash in 5 days.  Before, it was crashing after minutes of turning on the discreet GPU.

I also decided that IF i buy another macbook, il always be sure to sell it before applecare runs out. Its seems as if they're designed to last 3 years and then they all self destruct

Gapulus wrote:

 

Its seems as if they're designed to last 3 years and then they all self destruct

Ahh, so my 2009 MBP is 2 years late for its scheduled failure? (and what about the millions of Macs that are overdue for their self destruction date?

Who should I complain to about that?

Must be a manufacturing error if it's still going after 3 years. Must have slipped through the net. U might wana report it.

<Edited By Host>

I have 2010 as well as an 2009 and a 2011, they're all past the sell by date by now, how unreliable, they don't even fail when they are supposed to

Congratulations

You could try using the HSGT bootable utility to downgrade the SATA mode on the drive to SATA2 to check if its a handshaking problem - think it's called feature tool.  Not got access to my Mac Pro to double check exactly.  If the drive works fine when downgraded you will have to change the SATA cable to the logic board.

As far as other MacBooks having thermal issues - they do all improve thermally when I go contrary to Apple's guidelines by repasting and lapping the plate - but it's only our 15/17 inch 2011 models that needs this optimisation as a de-facto change to the engineering guidelines so they all fail far, far less.  The other models though I am not happy with the performance being good enough they are adequate enough to last longer than 3 years.

The 2012 models have a much smaller and cooler pair of CPU and GPU dies to cool and the later retina's have a hugely improved cooling system full stop compared to the palm tree layout in the earlier unibody designs.

It's not about self destruction. It's all about tradeoffs. If you choose a powerful machine in a sleek case, you have to accept a tighter thermal balance. If you want to be able to safely operate under tougher environment conditions (dust, heat etc.) you have to accept a larger size an a higher weight and give up some performance as well.

Mileage may vary but nothing lasts forever.

The problem is, Apple's after sales service is failing to meet our expectations. A "Pro" machine has to be able to withstand a "Pro" workload with no fuss. If it doesn't, Apple has to be on your side.

I'm afraid I wish I was as optimistic as yourself, but once you see artefacts the solder ball failing is inevitable though you may have delayed it somewhat.  Though if you ever think about re-doing it polish the plate for even better results, on the 2011 it is nearly as important to maximise the efficiency of the system getting that heat out quickly.

I've got machines in the field I repasted with AS-5 for years, first gen legacy MacBook/Pro that are stuck on Lion which is 7 years, PC and old gaming rigs etc etc.  I have two machines at home that I repasted over 5 years that idle temperatures are still excellent once you clear any dust out of the system which is the normal cause of long term heat build up.

alessiodd wrote:

 

It's not about self destruction. It's all about tradeoffs.

And I never said that it was, read again.

This chap (Gapulus) made that silly claim, you should reply to him instead of me.

hi Gav,

on the third replacement board, during the light use (e.g. playing video) temperature reads are about: CPU A 40°C & CPU Diode 45°C, GPU 45°C & GPU Diode 25°C and machine runs completely silent, as usual (I am using "temperature monitor" and "temperature gauge pro" to read all sensors, as smcfancontrol only controls the fans, I found...)

That looks ok now... I guess that for all those who used the computer for rather "light" applications over the past three years, computer will still run. However, after the extensive use of GPU hungry applications (e.g. longer gaming sessions), over the time same logic board shows higher temperature results even during the light use... unfortunately I did not keep temp readings of the previous board to compare, but I remember very well that a very first symptom of the strange behaviour with my two damaged logic boards was that during the rather light use fans produced constant loud noise... after while a "blue-striped screen of death" followed.

What is still strange to me is that the GPU Diode readings are always about 15 to 20°C lower compared to the GPU proximity temperature... and I noticed that fans are kicking-in with full 6200rpm only after 5 to 10 minutes being over 97°C. (I am planning to open it, polish the heat sink and put some high quality thermal compound.. in the meantime I emailed furmark burn-in temperature readings to the Apple reseller that made replacement and ask them to analyse, as Apple Care told me that 99°C is high, just to see how they respond...)

Csound1 wrote:

 

Gapulus wrote:

 

Is there anyone who has this problem that HASNT upgraded the RAM OR the hard drive? I had the exact same problems but then when i put my original hard drive and ram back in i couldnt get it to crash again to prove the problem to apple and so far it runs perfectly. Im a bit confused as im sure 7200rpm hdd and 8gb ram couldnt cause so much more heat than the standard components, or am i wrong?

Excessive heat is almost always caused by software issues, runaway processes.

Here, 'almost' is the magic word. Well-designed hardware dissipates heat through proper ventilation, and shuts itself down if this no longer works, unlike MacBook Pros which slowly melt down.

eezacque wrote:

 

Well-designed hardware dissipates heat through proper ventilation, and shuts itself down if this no longer works, unlike MacBook Pros which slowly melt down.

Really?

You can link to something that supports that I assume, it couldn't possibly be just one more opinion, could it?