Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I agree 100% and remember that not only Apple has this dilemma but all manufacturers who used the bad 8600m in them. I truly think this is the worst scam I have ever remembered seeing in the industry this widespread. Nvidia is ultimately going to be up the creek over this no matter what they do.

In the end the fiasco is nothing short of ridiculous.

So true. That's why Nvidia offered up the $200 million for repairs - that was getting off lightly. Still not sure why Apple did not dump Nvidia but continued to use them after they finally fessed up in July (Nvidia knew about it long before then).

Was it because Nvidia offered them such a great deal? Because otherwise, if Apple pulled out due to this mess, their name would have really been worthless.

But the bottom line for us, is no one is telling when the faulty 8600 gpus stopped. Having read somewhere that Nvidia was renaming the 8600s, my wild guess is that any gpu named 8600 (and as we know even some others) have the faulty soldering issue. Think about it - if you did fix the issue and continue to make the otherwise same gpu, wouldn't you rename it too?
 
Still not sure why Apple did not dump Nvidia but continued to use them after they finally fessed up in July (Nvidia knew about it long before then).
By that point, it was probably too late for the aluminum MacBooks to be redesigned. If they even could be--I don't think Intel or ATI has on-the-fly chipset switching yet.
 
By that point, it was probably too late for the aluminum MacBooks to be redesigned. If they even could be--I don't think Intel or ATI has on-the-fly chipset switching yet.

+1

You cannot redesign an entire motherboard's chipset in all of 7-8 months (when this was "discovered"). It simply ain't possible. That would have been an engineering nightmare, and a delayed macbook (pro) for sure.
 
I see. Thanks for explaining that.

When someone turns in an "early" 08 mbp (pre Oct 14) mpb where the 8600 gpu failed, what are they replacing them with?
 
So true. That's why Nvidia offered up the $200 million for repairs - that was getting off lightly. Still not sure why Apple did not dump Nvidia but continued to use them after they finally fessed up in July (Nvidia knew about it long before then).

Was it because Nvidia offered them such a great deal? Because otherwise, if Apple pulled out due to this mess, their name would have really been worthless.

But the bottom line for us, is no one is telling when the faulty 8600 gpus stopped. Having read somewhere that Nvidia was renaming the 8600s, my wild guess is that any gpu named 8600 (and as we know even some others) have the faulty soldering issue. Think about it - if you did fix the issue and continue to make the otherwise same gpu, wouldn't you rename it too?

So does the 9600, in all likelihood.
 
When someone turns in an "early" 08 mbp (pre Oct 14) mpb where the 8600 gpu failed, what are they replacing them with?
From reports here, boards that haven't failed yet, but aren't noticeably different otherwise. It may be that their chips had a different manufacturing process, but there's no publicly-known indicator of that.
 
My MBP is from the NVIDIA problem period and has now started occasionally glitching where a duplicate menubar and dock appear. i can't click them, they're just there. seems to disappear as randomly as it appears.

also random content someimtes appears on the desktop at the same time, like in this screengrab, where safari appeared on the desktop (a safari page from weeks ago as well?) but it was just the image, not safari running WTH?
(you can't see the second menubar so well in this one but it's there underneath, slightly offset)




Anyone see this before? Can't find any threads..
 
So if you buy a refurb 2.5GHz Classic MBP now with 512MB 8600GT will it have the newer board or still the ticking time bomb one?
 
(you can't see the second menubar so well in this one but it's there underneath, slightly offset)

Do you mean a second dock? I see a ghost of your dock, but not of the menu bar. That is indeed weird and probably counts as "scrambled video", but I'm not so sure. Here is a gallery of screen shots of machines with the GPU problem.

So if you buy a refurb 2.5GHz Classic MBP now with 512MB 8600GT will it have the newer board or still the ticking time bomb one?

It could have either. You wouldn't know unless you could look up the serial number to see when it was manufactured.
 
Do you mean a second dock? I see a ghost of your dock, but not of the menu bar. That is indeed weird and probably counts as "scrambled video", but I'm not so sure. Here is a gallery of screen shots of machines with the GPU problem.



It could have either. You wouldn't know unless you could look up the serial number to see when it was manufactured.

From what I can determine, it seems each and every 8600m manufactured, regardless of what notebook it was put into, is bad. MBP 2.4, 2.5 HP, Dell, etc...
 
Wishful thinking

All the 8600s are ticking time bombs. This desire to believe that some are somehow better is just wishful thinking.

Bottom line Apple has put us all on a two year rotation and will glady install a new ticking time bomb in to replace your existing one.

And my MBP was made in October but I'm under no illusion that it's any different. A request from Apple to explain "the difference" was an utter joke. After being placed on hold I was told of a software patch.

Pathetic.
 
And how do you think that Apple are going to do something else. Magically update an old design to replace the 8600. It's an NVIDIA problem, Apple just have to deal with the fall out!

Having said that I wonder what % of MBPs have had a problem, mine is fine so far and I expect if it does go they will keep replacing it from other posts and Apples press releases. Fine for me with an Apple store near by.
 
Do you mean a second dock? I see a ghost of your dock, but not of the menu bar. That is indeed weird and probably counts as "scrambled video", but I'm not so sure. Here is a gallery of screen shots of machines with the GPU problem.

There is a second menubar too, better seen in other screengrabs I can't be bothered uploading right now.

After the above post I considered it could be a Leopard bug but I can't find any other occurences of it, plus the weeks old Safari page appearing on screen is just plain weird.

My confidence in it is shot and I'm not looking forward to a full Erase and Install just on the off-chance something in my Install is corrupted but I guess I have no choice :(

I'll give it a week or two to see if it reoccurs or exhibits different symptoms.
 
If the problem still exists, then why on earth are Apple still selling it in the 17" MacBook Pros? All it is doing is costing them more money to fix them when they break. Also the 8600s must still be in manufacturing process because where will they get them from for the 17" MBPs? So they must have been fixed or Apple will have broken the contract between them already.
 
If the problem still exists, then why on earth are Apple still selling it in the 17" MacBook Pros? All it is doing is costing them more money to fix them when they break. Also the 8600s must still be in manufacturing process because where will they get them from for the 17" MBPs? So they must have been fixed or Apple will have broken the contract between them already.

Well, that's what any rational person would assume, but there are a couple of people on Apple Discussions and at least one here who had an October replacement GPU fail almost immediately. Of course that makes me wonder if the broken GPU didn't damage something in the computer that will make replacements be more likely to fail.
 
Of course that makes me wonder if the broken GPU didn't damage something in the computer that will make replacements be more likely to fail.
As long as they're replacing the entire logic board while doing the GPU replacements (which they more or less have to do--I doubt they're going to replace individual chips on the board!), that shouldn't be possible.
 
Well, that's what any rational person would assume, but there are a couple of people on Apple Discussions and at least one here who had an October replacement GPU fail almost immediately. Of course that makes me wonder if the broken GPU didn't damage something in the computer that will make replacements be more likely to fail.

He probably got a replacement GPU from the bad batch before the new ones were replaced in that store. I hope so anyway:confused:
 
There is a second menubar too, better seen in other screengrabs I can't be bothered uploading right now.

After the above post I considered it could be a Leopard bug but I can't find any other occurences of it, plus the weeks old Safari page appearing on screen is just plain weird.

My confidence in it is shot and I'm not looking forward to a full Erase and Install just on the off-chance something in my Install is corrupted but I guess I have no choice :(

I'll give it a week or two to see if it reoccurs or exhibits different symptoms.

I looked at your screen shot and it doesn't look like any of the failed GPU images I've seen before. There's definitely a problem but I just don't think it falls under this Nvidia umbrella. The ghost images coming back later on is weird, never heard of that before. Just in time for the Halloween season.

He probably got a replacement GPU from the bad batch before the new ones were replaced in that store. I hope so anyway:confused:

Keep in mind that every type of complex electronics like a GPU is going to have at least some failures. I've yet to hear any formal announcement about when Nvidia fixed the problem w/ the 8600s. They can't legally ship product they *know* are faulty and since there's been some announcments it can be proven they now know about it as does Apple. It was a manufacturing fault and to fix it they need to use a different solder material. I don't even know how far the admission goes, has Apple or Nvidia even admitted it was a faulty material? ...all I've heard the parties involved is rhetoric about 'bad batches' and that not all 8600s are faulty. Alot of tiptoeing around in a mindfield if you ask me. A class action lawsuit could still come down hard in this situation depending on whats going on here. You'd like to think that they *fixed* the problem and are replacing the faulty logic boards w/ GPUs that aren't vulnerable to the exact same problems. They would do the right thing wouldn't they? ;)
 
But the 17" MBPs have the 8600 in them, if they were faulty and both Apple and Nvidia know, there could well be a lot of law suites against them BOTH. Also what does w/ mean lol?
 
All the 8600s are ticking time bombs. This desire to believe that some are somehow better is just wishful thinking.

It's not an 8600 (as in chip) problem. It's a SOLDER problem. I've seen no studies or mention to indicate that all boards ever made on all computer systems out there that use 8600M GPUs all use the same solder. Nvidia says they don't all use the same solder. Some 3rd party said they're ALL defective. Not everyone who has this GPU has a problem. So it really comes down to whom do you believe?

Soldering is an odd thing. You can get bad solder joints in batches, on single chips or in this case, it sounds like the solder can crack under high temperatures. A software patch that cranks up the internal fan so the chip doesn't get so hot that the solder might fail COULD avoid the the problem entirely (given Nvidia says it's temperature related failures of the solder) for many people, so it's also disingenuous to scare people into not doing anything because it's a waste of time when it's not.

Yes, the situation sucks, but it doesn't help anyone to make it appear even worse than it is. You could have bought the new generation MBP and put up with a near mirror reflective glass screen (just saw them for the first time today and I'm glad I bought the old MBP while they're on sale) and one less firewire port.
 
What temperature does the solder become weak at?

From what I've been reading in the past 10 minutes in order to try and give you a specific answer, it's not so much one temperature but rather high-lead solder (which apparently is what Nvidia was using) is prone to cyclic failure, that is it weakens over time with large temperature changes. Laptops typically get powered up/down many more times than a desktop part, so this makes them more susceptible in general to such stresses. Increased fan output can help, but from what I've read, many believe that will only delay the problem, perhaps outside a warranty period so manufacturers don't have to take responsibility for the problem. Too bad the chips are likely surface-mount or one could resolder the chips relatively easily with higher grade stuff and then forget about it.

What I don't understand is why Nvidia, once they discovered the root of the problem, didn't immediately simply switch the solder they were using to lower lead or no-load solder and then newer batches would be without the problem (some comments from Nvidia seem to indicate they DID do that, but 3rd party articles poo poo everything Nvidia says and declares them ALL faulty, suggesting Nvidia preferred to HIDE the problem and keep on turning out bad batches. The question becomes which is true?) If the problem is in all the batches, Nvidia SHOULD be forced to offer higher quality solder replacements for all units, IMO. Extending a warranty will not fix the root problem. If, however, some batches did have better solder, they should release the information about which batches contain the fix. Otherwise, I would assume they're all bad since it makes no sense for them to hide the good batches.
 
But the 17" MBPs have the 8600 in them, if they were faulty and both Apple and Nvidia know, there could well be a lot of law suites against them BOTH. Also what does w/ mean lol?

with

From what I've been reading in the past 10 minutes in order to try and give you a specific answer, it's not so much one temperature but rather high-lead solder (which apparently is what Nvidia was using) is prone to cyclic failure, that is it weakens over time with large temperature changes. Laptops typically get powered up/down many more times than a desktop part, so this makes them more susceptible in general to such stresses. Increased fan output can help, but from what I've read, many believe that will only delay the problem, perhaps outside a warranty period so manufacturers don't have to take responsibility for the problem. Too bad the chips are likely surface-mount or one could resolder the chips relatively easily with higher grade stuff and then forget about it.

What I don't understand is why Nvidia, once they discovered the root of the problem, didn't immediately simply switch the solder they were using to lower lead or no-load solder and then newer batches would be without the problem (some comments from Nvidia seem to indicate they DID do that, but 3rd party articles poo poo everything Nvidia says and declares them ALL faulty, suggesting Nvidia preferred to HIDE the problem and keep on turning out bad batches. The question becomes which is true?) If the problem is in all the batches, Nvidia SHOULD be forced to offer higher quality solder replacements for all units, IMO. Extending a warranty will not fix the root problem. If, however, some batches did have better solder, they should release the information about which batches contain the fix. Otherwise, I would assume they're all bad since it makes no sense for them to hide the good batches.

Probably the most level headed post on this subject that I've read.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.