Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Nice of you to sound so eloquent but misread my entire argument in the past posts. Your assertion that "those that bash the dGPU are the same ones that have no business using the dGPU" is false. In fact, I speculate that even before this post, you had no ideas of the dGPU failures, as posts from owners have been sprinkled in a variety of Apple forums, and had Apple even post official repair programs. For you, in light of these, to still say, "hurr durr you people bashing dGPU's know nothing about it" is just childish. I did my own research, how about you try to do it next time?
I thought limiting my words would liberate you from drawing your erroneous conclusions, clearly I was mistaken.

To assume I am unaware of the dGPU failures is intellectually sloppy. I’ve been a proponent regarding those who’ve come across such failures since I joined the site. Where my opinion loses favorability with those such as yourself is that I think Apple has done as much as one can reasonably expect. throAU, tubeexperience and Queen6 could attest to this much. Perhaps take a page from their book and read what I am actually typing before twisting the dialog to suit your own agenda. I'll even direct you to said posts. Click Here.

Next, moving onto the next step of my argument, for you to suggest OP to invest in a large amount of money for dGPU models of MBP's in light of these failures that plague even the most recent models is silly. Your counterargument seems to be, "Well I know what I use MY MacBook Pro for, and you clearly don't need a dGPU." That's not the point. Regardless of usage of the dGPU (As I actually use it, even though it's a rebrand of the 2012 model, which is for shame), the point is is that Apple will NOT look to fix the dGPU problems associated with the Macbook Pro's. If you were to extend this argument and suggest a PC with a nice GPU, I would agree. But given the fact that Apple is so infatuated with a slimmer design at the cost of performance, choice of GPU, and even user experience, I don't see how you can ever suggest or even use a dGPU model of the MBP.

Just read what I wrote. I never suggested the OP purchase the dGPU. Time and again I’ve clearly proposed the iGPU being better suited for the OP with the information provided. I suggest you read the following excerpt from Hávamál, as you seem to not have a buffer between your brain and your keystrokes.

For the unwise man 'tis best to be mute
when he come amid the crowd,
for none is aware of his lack of wit
if he wastes not too many words;
for he who lacks wit shall never learn
though his words flow ne'er so fast.

And if you're trying to shove in the argument of, "Well I've never had this problem," don't bother. That circle jerk argument is so overdone and so useless in terms of seeing who's got what problems, and analyzing the situation, that it's hilarious that people still type in those comments.

Where did I say any of that? It's impertinent that you're directing said response towards me in your less than appropriate sentiment. Keep your ridiculous comments relative.
 
Last edited:
I thought limiting my words would liberate you from drawing your erroneous conclusions, clearly I was mistaken.

To assume I am unaware of the dGPU failures is intellectually sloppy. I’ve been a proponent regarding those who’ve come across such failures since I joined the site.../QUOTE]

The man has even wrote poetry. That's great. Let's look at your responses before you went full Shakespeare:

"Those that bash the dGPU are the same ones that have no business using the dGPU. As someone who uses FCPX, I wouldn't consider purchasing the MacBook Pro without dGPU. As far as battery is concerned, the dGPU actually maximizes my battery longevity by minimizing the amount of time the GPU is under load as it is able to accomplish tasks considerably faster than the Iris. When not using software that utilizes the dGPU, OS X, especially El Capitan has been smart enough to switch to the GPU that will use the least amount of power."

"You obviously don't need the dGPU. Save your keystrokes."

"In a language full of polysemy and homonymy such as English, opportunities for misreadings are bound to arise. So I will break it down to a point perhaps even you can understand. I need the dGPU. You clearly do not."

Here you're clearly trying to present yourself as the almighty master of MacBook Pro's. "Pft, you don't know what you're talking about. You don't NEED a dGPU. Because of my needs and how I use it, I KNOW about the dGPU and you don't. I would NEVER use a MBP without the dGPU."

Instead of trying to fluff up your nonsensical arguments with great vocabulary, how about this?

>Those that bash the dGPU are the same ones that have no business using the dGPU
- Or, we're users who've been afflicted with the dGPU problems that Apple has a horrible track record of, officially with the 2011-2013 MBP's and the 2013 MP. To portray yourself
 
Anecdotally, a huge number of our work machines fail around the same period (4-5 years), but because we depreciate them over 3 years (1 year if they are working on a remote site - mining industry), and we don't want to have to support a huge number of different models in our SOE platform (updating and testing our image for machines that are essentially worthless and due to fail is pointless), we budget on replacing them as soon as the extended warranty is out.

If a 4 year old machine comes in with a problem of any kind (for example) it is replaced. It's not worth fixing, when the hardware is out of warranty and at risk of failure again after we fix it (causing lost time for the employee, wasting IT time, etc.).

e.g., we fix it, send it back out... it fails again.
Say the lost time for the staff member is $300 worth of productivity (one day of disruption), and it costs IT half a day of screwing around (another $150 worth of productivity loss) you're almost halfway to a new machine worth of costs anyway. If its a machine for a VIP, double or triple that productivity loss $$ rate. 1-2 failures that could have been prevented by replacing the machine and you've blown any cost saving you thought you had by stretching out the life, even before you budget for the cost of replacement parts. To really save money based on the downtime impact, you need to have less than 1 additional failure within the 3-6 year period and it's just not going to happen that way.

And that scenario above is being optimistic. If the failure happens when the employee is on the road and they're disrupted for multiple days (e.g., until they get back, or due to IT backlog), it's even worse.

I'm sure our business is not alone in that.

I suspect that if enterprise was running 15" discrete GPU machines everywhere and tax law didn't make depreciating over 3 years and replacement at the end of that (or just leasing) attractive, then Apple wouldn't be singled out for GPU failure being a problem quite so much.

But Apple are often personal machines, they stay looking current for longer and people have higher expectations. But the insides are mostly the same...

I am in the energy industry (Oil & Gas) predominantly Exploration & Production, when I was in corporate the IT business model was pretty much the same with systems being written off at the 3-4 year point. Some of the engineering staff do use portables with dGPU out of necessitation, to me they never proved to be more of an issue than the basic notebooks, equally they are portable Workstation class costing more than double a 15" rMBP, some cases triple. As I was in operations, I do not have exact details on IT or figures to represent reliability.

These days I consult and buy my own hardware, which I swap out at the 24 month point, as this works for me. I have never incurred a hardware failure. The systems are pushed hard and pay for themselves in a very short period of time. I generally run with 2-3 systems with all data mirrored, I can switch between OS X & Windows ease, being OS agnostic.

As for Apple they know what they are doing and the likelihood of failure. My own background in highly complex drilling & measurement tools, employing highly sophisticated electronic packages alone tells me that the MBP design is compromised for the sake of form. Rapidly heating and cooling components such as the dGPU to the maximum limit is really asking for trouble from an electronic perspective as absolutely it induces thermal shock.

Even Mil Spec electronic chassis that are rated for up to 150C need the operating temperature raised under testing in a highly controlled manner and equally cooled down accordingly to lessen the risk of thermal shock and or catastrophic failure of the electronics, admittedly this is literally hundreds of K USD, potentially millions. Apple just uses the better end of consumer electronic components, certainly not Mil Spec or High temp rated.

For the video & audio professionals one way to tackle this is the keep the system under some form of load and the dGPU activated, as it`s the temperature delta & frequency of change that kills the chip. I still have an old afflicted 15" 2008 Classic MBP and it`s still running 100% stock, I believe the sole reason why the dGPU did not fail was this system was very rarely shutdown and always had the dGPU activated, I also took as many steps as practical to keep the dGPU below 80C -85C, frankly if Apple did the same there would be very little discussion regarding dGPU failure.

It`s also a little difficult to document easily as the effect is not linear, with the higher the temperatures and more rapid the heating & cooling the chances of component failure is significantly increased. I observe that some Windows OEM`s look to keep the dGPU below 85C, which makes a lot of sense at the component level. Apple clearly disregards this, a point worth noting.

Apple is not all bad they make product that people really like and can engage with, equally the the MBP with dGPU is simply best avoided, even if you have a need for the dGPU there are far better solutions. Those will say there is no issue, reality is that they likely do not need the dGPU as an integral part of their workflow, so the the chip is never pushed.

My own opinion is that if you need or want a portable Mac with dGPU fine, just be aware that it will likely fail sooner, rather than later. If like myself you can monetarize the hardware who cares, just swap it out in good time, I pass my old systems to family or a good cause.

Q-6
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.