According to my discussion with Apple today, the long-standing flashing / scrolling graphics defect present in thousands of MacBook Pros includes ALL machines produced with the NVIDIA 8600GT chip, as well as the unibody MacBook Pros with the NVIDIA 9600MGT / 9400M chip.
I first reported the problem a few days after getting my MacBook Pro 17" in March of 2008. That's nearly TEN MONTHS AGO!
At first, it appeared this defect might be related to the defective NVIDIA chip problem. But I've had two motherboards swapped out, and the defect remains.
The defect typically shows up AFTER users have woken from SLEEP mode when on battery. It is triggered for some users by a combination of Quartz-enabled applications and trackpad operations, such as two finger scrolling.
There's a relatively easy way to see if this is the defect you're experiencing: Using the FREE "Quartz Debug" utility, available as part of Apple's Xcode download, select TOOLS > DISABLE QUARTZ EXTREME. The defects will go away, along with any speedy graphics performance you might have been enjoying. Once Quartz Extreme is disabled, try dragging a window with Quartz Compositing (like a Finder window) and you'll find it barely moves.
Restarting your computer will typically correct the problem -- for a while.
Hoping to find out if there was ANY MacBook Pro hardware immune to the problem, I queried AppleCare product specialists today, who consulted MacBook engineering. They indicated that ALL MacBook Pros from 2008 THROUGH 2009, including the new Unibody MacBook Pros, have this defect. They said the defect was related to the NVIDIA firmware, which was the same firmware used in the 8600GT, 9600MGT, and 9400M.
Why some users are affected and other aren't has been a long-standing mystery.
For a brief period, OS X 10.5.5 APPEARED to "correct" the bug, but for some users the bug did not go away. It has been universally reported (for example on Apple's Discussion boards) that 10.5.6 BROKE this particular "fix", causing thousands of users to suddenly see the defect again.
In the end, I'm confident Apple will "get it right" -- they usually do. But ten months is an awfully long time to go with such a significant defect.
Anyone else have any information on this issue?
I first reported the problem a few days after getting my MacBook Pro 17" in March of 2008. That's nearly TEN MONTHS AGO!
At first, it appeared this defect might be related to the defective NVIDIA chip problem. But I've had two motherboards swapped out, and the defect remains.
The defect typically shows up AFTER users have woken from SLEEP mode when on battery. It is triggered for some users by a combination of Quartz-enabled applications and trackpad operations, such as two finger scrolling.
There's a relatively easy way to see if this is the defect you're experiencing: Using the FREE "Quartz Debug" utility, available as part of Apple's Xcode download, select TOOLS > DISABLE QUARTZ EXTREME. The defects will go away, along with any speedy graphics performance you might have been enjoying. Once Quartz Extreme is disabled, try dragging a window with Quartz Compositing (like a Finder window) and you'll find it barely moves.
Restarting your computer will typically correct the problem -- for a while.
Hoping to find out if there was ANY MacBook Pro hardware immune to the problem, I queried AppleCare product specialists today, who consulted MacBook engineering. They indicated that ALL MacBook Pros from 2008 THROUGH 2009, including the new Unibody MacBook Pros, have this defect. They said the defect was related to the NVIDIA firmware, which was the same firmware used in the 8600GT, 9600MGT, and 9400M.
Why some users are affected and other aren't has been a long-standing mystery.
For a brief period, OS X 10.5.5 APPEARED to "correct" the bug, but for some users the bug did not go away. It has been universally reported (for example on Apple's Discussion boards) that 10.5.6 BROKE this particular "fix", causing thousands of users to suddenly see the defect again.
In the end, I'm confident Apple will "get it right" -- they usually do. But ten months is an awfully long time to go with such a significant defect.
Anyone else have any information on this issue?