Mine was repaired under the program, and it's running COOL and FAST.
I use gfxCardStatus and keep it in integrated mode just to be on the safe side, and there's no sign of trouble - its a bit of a pain that you need to switch it every time you boot up.
https://www.macrumors.com/2012/07/1...s-easy-control-of-macbook-pro-graphics-cards/
https://gfx.io/
If it failed again I'd have it re-balled with new GPU and lead based solder.
The two 17" MBP's we have in our home will be with us for a long time.
My 2011 has a 1TB SSD, USB-3 Express card, and the Airport card to give me the Continuity features in OS X 10.10+, including Handoff, Instant Hotspot, and AirDrop
http://www.aliexpress.com/item/Orig...-A1286-2011-2012/32397072045.html#60677026420
A good advice!
I will add this to my other preventive action. Thank you.
In our family and friends around me the are several 15" MBPs... And aside of the infamous GPU-issue, everybody is satiesfied. Everyone went to a very good Apple accredited service here and led run the tests. So far they passed the test and work well *knocking on wood*... Until december 2016 we are now covered as the "extended repair program" got another extension - and after that it costs about 200 EUR or USD to let repair them by a SERIOUS 3rd party service.
BTW: it seems that there is also a problem with a crappy capacitator near the GPU localized just on the othervside of the main board...
As for PREVENTIVE Action to lower the risk:
I stated already some weeks ago how to get a little bit more on the safe side:
Just my general advice for prevention of GPU failure:
In our family we have several 2011 and 2012 MBPs. Although ther were some OTHERS than gpu failures with them, i did some prevention for GPU issues, hoping they will have sort of effect on the risk for it...
I am convinced mostly these issues arrive more likely with overheated GPUs and possibly oberheated less known components (see above: capacitators etc) .
My hypothesis: "Overheating" begins already at temperatures beneath the official limit, there is not a threshold seperating working condition that have zero risk from others that have extremely high risk, the fubction of risk is instead increasing more and more with temperature - maybe because of the ROHS issue and the ridiculous minimizing of fan action and "thinness-fetish" of Jony Ive. It seems also that the fan is reacting too late on rising temperatures for "comfort" reasons (noise from cooling fans). There are wrong priority settings (function follows form, fashion and "silence" instead of form follows function and reliability).
So my personal solution is:
1). To Prevent absolutely impaired passive convection and use the MBP the less possible on a cushion or in other positions that stop or minimize convection, especially the little part under the CPU/GPU which has just the size of a postcard. Not only the GPU itself, but as well the CPU will heat up the inner part of the MBP and so also the GPU secondarily. Sadly, the back cover of the "unibody" model is far to thin to direct local heating to other parts of the backside, as you can easily feel with your hands 90% of passive convection is restricted on a very little area of not more than about 6x8 cm just beneath CPU and GPU. This might also indicate insufficient design of the active part of the cooling system, like laggy reaction on rising workload and temperatures. Interesingly, the fans are NOT in first place reacting on temperature, but on workload, which is smart. Because theoretically it acts more in a preventive than reactive way - BUT the way the fanspeed is programmed to react on workload is insufficient. The problem is that the "good idea" of workload as a trigger for fanspeed gets contraproductive when simple placement of the machine and so passive cooling part gets more important for the actually temperatur than electrical workload. A firmware modification that respects that would be better: rising fanspeed EITHER if electrical CPU/GPU workload rises OR temperatures of CPU/GPU rises above a certain temperature (let us say beginning smoothly above 55-60 degree celsius with still silent augmentation from standard idle mode (=2000 rpm) to 2,500-3,000).
2) installing apps like MacFanControl:
a) to diagnose easily with just one click the temperatures of CPu and GPU. With MFC you have also the option to permanently have the temperature of the dGPU being shown in the upper bar of the screen, and also the rpm of the cooling- Very nice because like that you have everytime diagnostic control of the temperature in a fraction of a second!
b) to react sooner than the laggy system itself on temperatures above of 55-60 degrees Celsius by manually choose/adjust higher rpm of both fans. This also takes just some clicks and about 5-10 seconds at max.
Normally they run at about 2000 rpm - nevertheless even 3000 or 3,500 rpm will cause nearly no significant noise - but significant lower temperature!
3) using this configuration you can also already choose modestly higher rpms
before You charge your MBP with high workload for CPU and/or GPU using MacFanControl.
4)it makes you LEARN about cooling management:
Regarding the temperatures in using Fan control apps makes you better understand how easily you expose your mBP to damn high temperatures (due to the rather mediocre cooling design of the MBP, and not due to wrongly using your machine!) . And so you learn to think about manual prevention of higher temperatures very easily.
3) CLEAN smoothly your MBP inside !
(Edit)
Under normal conditions, the MBP will saty pretty vlean inside - but over the years it might be an good idea to open the backside and clean it CAREFULLY - especially the fans themselfes and the open channels for air circulation of the unibody housing between the screen and the keyboard. Sadly even a thin laEr of dust is a good isolator of heat - so taking off dust may lead to better passive convection either. But be careful... Not too much air pressure nor suction and never wet papers or so... A soft little NEW brush with natural fibers may help better than any textile or cleaning papers...
This will surely not prevent ALL possible existing risks, bit at least some of them, I am convinced about this.
-------------------
Some general words about Quality management and Customer relationship:
The GPU-issue is clearly the result of very POOR engineering design - sadly, this is repetivly done over many years and models (since 2008 until 2013- rMBP model, that makes 7 years) and apple is seemingly unwilling or incapable of learning their lessons, as the "trash-can-style" new MacPro has now the same problem which was unknown with the super-well designed classic MacPro line before. Like already before with the MBP isssues over SEVEN YEARS the classic "reflexe" of the apple management about the GPU-card issue of the "trash-can-style MacPro" is very childish in denying of the problem for at least 2 yeras (evidently they either have no quality system established or they are incapable of using it - or they just give a **** about customers needs and evidently weak hardware engineering/firmware design) and then react just because class action lawsuit is done or near a negative judgement...
It seems that apple has the wrong priorities as well for hardware design itself as for a far too ignorant setting of the cooling fan reaction/temperature prevention.
Focussing on thinness and silent performance might be tempting - but only as far as the reliablility and stability of performance is guaranteed. They forgot a long time ago that the word "PRO" in their naming of some of their product lines means PROFESSIONAL, not "chic" or "nice to look at"... They transformed their pro-line into a fashion-line. "Thanks" to a hardware-designer who created once the highly appreciated design-ikone classic MacPro (perfect servicability and upgradability, meaning REALLY PRO) - and after that became more and more nuts or let's say absolutistic about "thinness" and now giving a **** on customers needs and ergonomy. The last cry of his more and more failing bizarre ideas being the "new" magic mouse with a charging connector UNDER (!) the mouse instead of putting it in the same place as in the Magic Pad2: in the same side as before to enable charging and using at the same time.... This is not decent design or engineering, this is real-world comedy! Just like the "one-connector-only" MBP...
As for the argument of some extreme apple-loyalists apple having "already done" sufficient action:
Nearly ALL "extended repair programs" started only after YEARS of denying and ignoring the design flaws. And even worse: often enough they risk class lawsuit action - and ONLY at the moment that they are CLEARLY going to lose the case they accept a "deal" of starting the so-called "extended rfpair program" ... SHAME on them !!!! Because that means that tens of thousands of loyal customers were hold in an expensive and quite embarrassing state pf incertainty both about costs and - more important - even reliability of their own machines!
And sadly, it began already long before Steve Jobs antenna-gate: when he took tens of millions of customers for fools claiming "You are holding it wrong" , the GPU-gate of the 2007-2009 MBP line had already taken place...
I switched from the windows platform to apple in 2008 and paid a super-prome price for a top-MBP 15" - and learned the hard way about "reliability" of the "Pro"- MB-line... After about not even 3 years my MBP was DEAD... Of course, apple denied at that time any responsibility... And like many others I burned nearaly 3000 EUR ... A damn expensive and hard way to learn about apple's reliablility and so-called "customer service"...
