Original Batch of Santa Rosa MBPs Failing

in Current Mac Hardware edited January 2014
As per http://apad.sytes.net/blog/

Recently, an unknown number of first-revision Santa Rosa MacBook Pros began exhibiting issues with their onboard video cards. After a reboot, or on wake from sleep, the machine refuses to acknowledge the presence of a display, either internal or external. From that point on, the computer never regains its displays - not after a reboot, etc. Subsequent debugging indicates that the machine is misidentifying its NVIDIA GeForce 8600M GT card as the MacBook?s Intel X3100 card. This issue is known to affect at least 50 people - a group of affected users has formed a Google Spreadsheet in order to document and organize cases.

AppleCare is recommending replacing the logic board, which some have gone through with, only to have the machine return to an unusable state shortly afterwards. Compounding the issue is the fact that this problem has arisen only shortly after the expiration of the default warrantee on these machines (this issue seems to only affect machines shipped around June, 2007), thus causing the logic board replacement to cost upwards of 400$ for those who did not purchase extended warrantees. If you are experiencing the issues detailed below, please add yourself to the spreadsheet and visit our thread on the Apple Support forums, so we can get a reasonably accurate count of affected users.

Symptoms & Notes

Blank screen, both on the internal and external displays

The computer boots; it is accessible over the network or with Screen Sharing

Target Disk Mode works (this can be used to backup user data!)

ioreg and System Profiler both report an Intel X3100 video card, which is incorrect

Affects people even with MacBook Pro EFI Firmware Update 1.5.1 installed (which was released to fix a similar problem, introduced in Firmware 1.5)

Seems to be independent of any software updates, hardware changes, etc.

There are a great number of potential fixes floating around on the Apple Support forum thread. These include PRAM/NVRAM/PMU resets, firmware restores, changes in memory configuration, deleting Safe Sleep files, etc. However, none of these appear to be permanent fixes; apparently, even replacing the logic board is not a permanent fix, at least for some!


  • Reply 1 of 4
    kennywrxkennywrx Posts: 141member
    hmmm... I wonder if it has something to with this... Defective Nvidia GPUs
  • Reply 2 of 4
    futurepastnowfuturepastnow Posts: 1,772member
    Wonder if it's the same defect that plagued the old iBooks. In them, repeated heating and cooling of the graphics chip weakened the substandard solder that held the chip in, eventually breaking its connections. The MBP's chips get a lot hotter, too.

    The fact that it's detecting an X3100 also makes sense, as the computers use the GM965 northbridge, which has the integrated graphics core. If it loses the GPU, it defaults to the IGP, but in the Pro it isn't connected to the display.

    Having typed all that... this is probably something that affects an insignificant number of people. Yawn.
  • Reply 3 of 4
    nvidia2008nvidia2008 Posts: 9,262member
    I sympathise and hope nVidia cleans up its act to stay ahead of Intel and AMD-ATI in its own way. It is very aggressive at new launches and acquisiton, hopefully their fabbing will improve, once they get to most of their GPUs with physics, cuda, massive-multicore, etc, at 45nm, hope things will be better.
  • Reply 4 of 4
    nvidia2008nvidia2008 Posts: 9,262member
    My nVidia 8500 GTs , two of them, have ben running well in SLI (PC) for about a year now. It survived a flight (not in original packaging) from South East Asia to London.

    They are factory overclocked (Gigabyte and Xpert Vision), and then manually overclocked some more.

    I suggest to all that have their MBPs recently repaired, as soon as you get it back, run an intensive 3D game that pegs the CPU and GPU at 100% for at least 6-8 hours to test if everything is okay.
Sign In or Register to comment.