MacBook Pro graphics chips failing prematurely

MacBook Pro graphics chips failing prematurely

Summary: It appears that the current crop of MacBook Pros seem to be suffering from a premature failure of their Nvidia graphics processors. According to a piece on The Inquirer "All Nvidia G84 and G86s are bad" and MacBooks Pro's with either the Nvidia 8600GT GPUs are prone to failure:The short story is that all the G84 and G86 parts are bad.


MacBook Pro graphics chips failing prematurelyIt appears that the current crop of MacBook Pros seem to be suffering from a premature failure of their Nvidia graphics processors. According to a piece on The Inquirer "All Nvidia G84 and G86s are bad" and MacBooks Pro's with either the Nvidia 8600GT GPUs are prone to failure:

The short story is that all the G84 and G86 parts are bad. Period. No exceptions. All of them, mobile and desktop, use the exact same ASIC, so expect them to go south in inordinate numbers as well. There are caveats however, and we will detail those in a bit.

Both of these ASICs have a rather terminal problem with unnamed substrate or bumping material, and it is heat related. If you ask Nvidia officially, you will get no reason why this happened, and no list of parts affected, we tried. Unofficially, they will blame everyone under the sun, and trash their suppliers in very colourful language.

The symptoms are no video output but the computer still boots. After the problem occurs and you attach an extrnal monitor the system profiler lists the graphics card as an Intel X3100, instead of the GeForce 8600M GT that it should (pictured).

It's difficult to say what Apple will do, they've extended warranties for dud batteries, so its conceivable that they would something similar for this but they haven't yet acknowledged the problem, so who knows.

The Apple Discussion Forums are buzzing with the issue and  someone has started a spreadsheet for affected users here.

(Tip: Eliot)

Topics: Processors, Apple, Hardware

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Wouldn't this imply that ....

    ... Apple is using the same faulty parts as Dell and HP and not the superior parts that Laff and others claim?
    • Apple hardware nowhere near what they are cracked up to be.

      I've read numerous complaints, reports, and blogs by Mac users about the notebook parts failing. That falls 100% responsibility to Apple. They are the ones who made and approved the designs. Apple approved all the parts.

      I have owned notebooks from Toshiba, Fujitsu, and Dell that have lasted much longer than Macbook's and Macbook Pros. I have 9 year old Fujitsu who's thats still working. The only failure was the HDD, but that was after 6 years of use. Constant reports of hard drive failures under the 2 year mark, needing to replace proprietary pieces that fail too often, and less than honest disclosure is why Mac products are banned from our household. Seeing the grief from our friends was enough to make us better consumers.

      Apple products are simply overpriced pure and simple. Why can't you guys get this through your heads? Even Tom's Hardware got it wrong. These articles continue to debunk the Apple superior hardware debacle. Go buy yourself an HP and save yourself some money.
      Solid Jedi Knight
      • *DING* *DING* *DING* We have a winner!

        Never mind Apple is an empty shell; going from their own OS with a superior hardware technology, to piddling IDE drives, Intel processors (Motorola had the better chips), and from their own OS to borrowing some open source for OS X... yet prices never came down to match...

        I know, all that extra money is going to environmental causes, yes? (Can't be; so many entities, right down to greenpeace, keep harping on this company for remaining in the dark ages...)
      • Amen & AMEN

        I started to buy into the Apple mindset 2 years ago & since I am going to school for electronic media I went ahead & bought a Macbook (& since most EMP people are Mac fanboys I will probably be stuck using them to a degree). Man, talk about overpriced. Then I get it & find it very un-customizable, screen flickered like mad, freeze up & a host of other things that led me to understand: Macs are just computers, they fail too. What sucks is that people are led to believe they don't fail. Then, since they have bought into the mantra, drank the kool-aid, spent too much & don't want to loose face, they go on about the service when their Macs do fail. Wait, your mac failed, its not supposed to. You paid way too much for a computer that uses off the shelf hardware, an OS which is mostly not done by Apple & isn't exactly great (though much better than anything Apple did themselves) & some pretty, cheap plastic housing... get mad @ them! I've had little problem with most of my PC's & what I do have its common knowledge, no spouting "It just works" like a parrot. Seriously, Mac fanbois, learn Linux & see what a real OS can do for free compared to your OS of choice.
  • How can this be?

    Just yesterday we were being told how Apple used better quality parts in their systems.

    Though since the source is The Inquirer I'll take this with a grain of salt.
    • The Expense Argument

      In the end, we're back to an expense argument. You had
      asked earlier for some evidence of Apple's TCO being
      cheaper. below is a link and a excerpt.

      You folks should be picking your battles more carefully
      and the arguments should be a little better crafted. They
      are coming across as petty envy.

      The clarion call "show me the facts" is a tacit admission
      you don't have them already. What's up with that? The exhalation of ignorance is hardly a good way to start an

      "The results of this TCO astounded me. For my small
      enterprise, owning a WinTel box for three years costs twice
      as much as owning a MacTel. When I talked with several of
      our clients, I found that the burdened cost of ownership
      per PC - just for support - ranged from $1,300 to $4,000
      per year."
      Harry Bardal
      • I asked for credible facts.

        This TCO study was debunk quite some time ago (the article is date 2/27/2006). It's nothing more than another Mac convert trying to justify his choice.

        Perhaps next you'd care to post a link to the Pfeiffer Consulting TCO "study"?
        • Debunking

          Surely you'd be happy to point me to that debunking
          then... link hound that you are.

          Sorry you can't accept the findings. They certainly aren't
          scientific and I'd be the first to admit it. Too bad, the guy
          seemed like he had some integrity. No matter, you're going
          to provide me with hard debunking data and we'll put it to

          And imagine, a two year old link. So much has changed
          since then. There's been Vista, and Vista SP1, and Vista SP2
          otherwise known as XP. Clearly the PC economy has made

          I've given you a link to a more topical, and equally
          "unbelievable" anecdote from ZDNet's Christopher Dawson
          within this thread.
          Harry Bardal
          • Then they're worthless wrt to my request.

            [i]Sorry you can't accept the findings. [b]They certainly aren't scientific and I'd be the first to admit it.[/b][/i]

            I want something other than someone's opinion that Macs are more productive. If true shouldn't there be mountains of support for this?
          • Where's the Debunking?

            Sorry, I was expecting evidence of debunking. Since that
            evidence is lacking it must not exist, therefore it is not

            Wait a sec, this is a logical syllogism and it's far beneath
            me and should be for you as well. I'm taking you at your
            word. I'll choose to believe you if you say there was some.
            In the absence of "proof" all we have is opinion and our
            ability to convince and that's all that's available to this
            forum. What did you think this was, a court of law? Why
            did you offer "debunking" if you were not prepared to
            introduce it?

            While you're trying to figure out who has the burden of
            proof, people are switching. Your contention that there
            should be mounds of impartial evidence for one side
            ignores the fact that there would also be mounds for the
            other if they were ahead. It's simply misguided.

            This "experiment" uses an newly open market as the pitri
            dish, its results are measured in switchers, and the unit of
            measure is a single autonomous individual with tech
            purchasing power. Sadly, proof of Apple growth need not
            equate to better TCO. Going forward, there will be nothing
            to compel the switcher to convince you. It's clearly an
            uphill effort, and there is nothing in it for them. They have
            done their own homework. You can rest assured that
            reactionary viewpoints may stay insulated from "proof" for
            some time.
            Harry Bardal
          • For starters right here:

            "They certainly aren't scientific and I'd be the first to admit it."


            You're welcome.
          • Congratulations You (I) Win

            If it walks like a duck, and if it quacks like a duck, DNA
            tests are clearly still required before a duck certificate can
            be issued. Ok, suit yourself.

            Because it's a great deal of effort, and an expense meeting
            your burden of proof, and because even the arguments
            against me have to come from me, we're going to expect
            you to quantify your requirements in a 47 page white
            paper. Single spaced please. Double spaces will cause it to
            be rejected and the contained information to be
            considered false. If, on some bright and shining day, these
            requirements are met, we'll be sure to keep quiet until
            such time as the data has been reviewed by you, and has
            been verified. We won't corrupt the process with opinion,
            observation, personal experience, or any other tainted
            input. All we'll do, is set your own criteria against your
            own certified result. On that bright and shining day, we'll
            expect a concession.

            You see what I'm doing here? I'm making sure that your
            currently obtuse opinion pales by comparison to how
            obtuse it will have to be in future if you stick with your
            current trajectory. My record of predictions have been
            astoundingly good if I do say so myself. I'm setting it
            against yours. Oh but it's flakey, I know, so if we have
            failed to prove, the burden shifts to you. In the mean time,
            you needn't participate in "opinion" oriented discourse, it's
            clearly too nebulous for you.

            Currently, the anecdotal TCO that has been returned so
            far, the one I choose to believe, requires that Gateway pay
            me between $900 and $3350 to "buy" their computer. The
            very one you say is a bargain. For me, you bet this issue is
            Harry Bardal
          • @ Harry Bardal: Not really.

            [i]You see what I'm doing here?[/i]

            When you started prattling on about duck DNA I quickly lost interest and just skimmed the rest of your post looking for something worthwhile to respond to. Needless to say this one sentence was it.
      • Or buy a PC and shove Linux into it.

        Save the extra money and buy 2x as many computers because of it.
    • Here's Another

      Here's another more topical opinion from ZDNet's own
      Christopher Dawson. Sorry, no hard figures for you in this
      one, but evidence is building. One has to wonder why any
      sensible person would pay out more in capital costs if they
      thought it returned less or the same value and productivity
      as a lower priced item? These "studies" albeit informal
      studies use real money after all. It's likely the same color
      and value as the money you may use. It's not going out in
      insignificant amounts. Refute the conclusions if you must,
      just please don't ignore the conviction.

      If the argument is "cultism" you're going to have to admit a
      growing number of members. With that comes decreasing
      chances of "cultism" being the reason. This is fantastic, we
      can actually have a dialogue about the meaning of metrics
      in the abstract without demanding them at every turn.

      I, for example, could now ask you to provide some
      examples that refute the ones I've given you. I won't
      though. A good argument has always been good enough
      for me. I have yet to see one from you, but I'll let you
      Harry Bardal
      • Then it's worthless wrt to my request.

        [i]Sorry, [b]no hard figures[/b] for you in this one, but evidence is building[/i]

        The only "evidence" building is that someone decided to chose Macs. If I were to point you to a place that chose Windows would this convince you Windows has lower TCO and better productivity?
        • We Have to Break Up

          Sorry, this just isn't working out. Having a desire to know
          what a rigorous scientific study might conclude is a fair
          and noble thing. I commend you for it. The lack of this
          "proof of theory" however should not short circuit intellectual curiosity. Indeed one of the pillars of good
          science is careful observation, and the formulation of that
          theory. Theories have to exist before they are proven. I'm
          offering one. You're rejecting it.

          Here's the funny bit, it's not your or anyone's place to
          dismiss a theory for the lack of an experiment. Take a
          lesson from Microsoft's marketing campaign, we had not
          lived on a flat earth until it was proven round, but any mediaeval surf who looked at the crescent of the moon
          could have gotten the idea.

          It's this obtuse nature that puts the Windows ecosystem
          into denial over what's going on. It's what's causing the
          overreactions to a limited Apple graphics card problem on
          the same day the impregnable Vista was breached. It
          causes demands for hard evidence and proof, when there
          will be none that is acceptable.

          This is ego, willful ignorance, and cognitive dissonance.
          What the freeking heck do you have to say about the stuff
          that falls outside the purview of hard science? If it is to be
          "nothing", and the reaction is to be "nothing", know that it
          will effect you regardless.
          Harry Bardal
          • You're got that right.

            [i]Sorry, this just isn't working out.[/i]

            Still trying to figure out why you bothered providing information you knew would not be accepted. Here, let me give you my TCO study:

            - Gateway PC with Vista Home Premium: $650

            That's it. No anti-malware software installed, no maintenence software installed (or performed). Just sit down and use.

            Did that convince you Windows has a lower TCO? Great. I'm glad we settled this difference.
          • Message has been deleted.

            Linux User 147560
          • I guess some of us are smarter than others.

            [i]Anyone with half a brain knows you can't run Windows safely without some add-on protection and expect it to stay clean and safe on the internet.[/i]

            Yet here I am doing it just fine.