Nvidia chips failing?

What nobody knows at this time, and what Rod is referring to, is there is no evidence that every nVidia GPU in every MacBook Pro will fail, and in fact I don't believe Dell has said that. A blogger has said that he believes every nVidia GPU is bad, but he has nothing at this time to back up that claim.

As this is not just an Apple issue and involves several companies it will take time to play out, just like the Sony battery issue did. But as Rod said, and I can also attest to, if you had the rate of failure that some people are claiming you would be seeing hundreds and hundreds of posts here listing the symptoms you have provided, and it just is not happening yet. Will it in the future? Who knows. Have GPUs failed on MBPs? Yes, but I also had an ATI fail on one of my desktop machines, with similar symptoms. When I checked this site I found out that some other people had the same problem, took it in for service, and got it fixed. Does this mean Apple had been "hit" by this? Hit by what?

If nVidia does determine they are all flawed, or Apple, HP and Dell determine they are all flawed, you can expect that some sort of financial agreement will be brokered and some kind of remedy provided. It is my understanding that all Dell is doing at this time is providing a driver that ups the fan speeds to keep the GPU cooler.

If you have proof that all nVidia 8600M GT GPUs are bad post it here. Until then people need to keep a level head and have some patience.

I have absolutely no connection with Apple whatsoever, doctorjo5 and neither do any members of my family. I just don't like seeing very dubious sources being used to "over-hype" issues like this. As the ars technica article that I mentioned near the beginning of this thread says, the only reasonable course in a situation like this is to "wait and see".

NVidia, as I also indicated, have said quite specifically that the issue only affects some chips and only occurs in some system configurations:

+Only a very small percentage of the notebook chips that have shipped are potentially affected, and the problem depends on a combination of environmental conditions, configuration, and usage model.+ (see http://arstechnica.com/news.ars/post/20080716-nvidia-denies-rumors-of-mass-gpu-f ailures.html )

and

+the failure appears related to the combination of the interaction between the chip material set and system design+ (see http://www.eetimes.com/news/semi/showArticle.jhtml?articleID=208802249 )

The specific "system designs" which have so far been shown to suffer as a result are made by Dell and HP. Is Apple also affected? Well, we simply have no evidence that they are. Apple have their own thermal system control routines and neither NVidia nor Apple have indicated that a problem exists with their own particular configuration. Nor can I see a higher than usual number of reports of graphics card problems on these boards (there are some, to be sure, but no more than usual as far as I can see) .

So, as I said, don't panic (and don't create panic in others). When similar issues HAVE affected Apple in the past, (with for example the very widespread capacitor issues that caused problems on a broad range of computers from many different manufacturers in 2004/5) Apple introduced an "extension repair" program once the nature of the problem had become clear. I see absolutely no reason to suspect that they will behave any less properly here, IF the issue turns out to be a genuine one.

Cheers

Rod

hello,

i never said that ALL nvidia chips are failing. i don't believe that dell has said that either. and i don’t presume that dell would speak for apple nor did i suggest this. so please, don’t put words in my mouth.

there is obviously an issue with the chips, and i would be HIGHLY surprised if apple has escaped this.

as to your suggestion that there would be hundreds of posts here if there was some sort of epidemic chip failure, well, not necessarily. that would only happen if there was a mass simultaneously failure of chips. the chips may fail incrementally. there is also the fact that people might not be aware of what the issue is and still going through apple support and like me, being told that their computer is "fine".

also, it does not appear to be clear what all the symptoms of the chip failure are. obviously, more information will be revealed as time goes on. anyway, here is a bit more info about the issue:

“ Symptoms include a lack of video output, a non-booting system, and also seemingly unrelated issues like a dead Wi-Fi adapter and a wonky battery charge light.”

interestingly the article says HP has offered an extended warranty for affected customers as well as a BIOS update, also HP began addressing the issue in november 2007. you can read the full article here: http://techreport.com/discussions.x/15219

here is a list of symptoms according to HP:

Step 2: Identify the symptoms

If your computer’s product number is listed above, and the notebook PC experiences one or more of the symptoms listed below, contact HP during the Duration of the HP Limited Warranty Service Enhancement to determine whether you are eligible for a free repair.

The following symptoms apply to Pavilion dv2000 and Presario v3000 notebooks:
• The notebook does not detect wireless networks and the wireless adapter is not detected in the Device Manager.
• There is no video on the computer LCD panel or external monitor.
The following symptoms apply to the dv6000, dv9000 and v6000 series notebooks:
• The notebook does not detect wireless networks and the wireless adapter is not detected in the Device Manager.
• There is no video on the computer LCD panel or external monitor.
• The notebook has no power and no active LEDs.
• The notebook does not start.
• The battery charge indicator light does not turn on when the battery is installed and the AC adapter is connected.
• The notebook issues a single beep during boot indicating no power.
• The external monitor functions but there is no image on the notebook LCD panel.

here is the link to the HP site if anyone is interested in reading more:

http://h10025.www1.hp.com/ewfrf/wc/document?docname=c01087277&lc=en&cc=us&dlc=en &product=1842189&lang=en#

and to be fair no one is creating a "panic" about the situation. personally, i am just a concerned apple customer who is still trying to understand what the problem is, and hopefully help others.

regards,

polisciguy

hello,

my pleasure. i hope you get your issues sorted there too. "the fact that staff are called 'genius' does not mean that the customer should be treated as 'idiot". LOL thanks i needed a good laugh. ya, i sometimes have a hard time coming to grips with what apple tech support tells me too. you would think i never used a computer in my life.

post back and let me know how far you get with apple. if i have any luck i will let you know too. hope it never comes to what you said in your first message.

regards,

polisci

"i never said that ALL nvidia chips are failing."

"there is obviously an issue with the chips, and i would be HIGHLY surprised if apple has escaped this."

Your second statement contradicts the first one.

One important point that has largely been ignored when this issue is discussed is that while the GPU's themselves may (or may not) all be the same, the machines in which they are installed most certainly aren't.

Looking over at the Dell forums on such matters, for example, I see that some people are reporting GPU temps of over 100ºC. After the BIOS updates many people seem to be pleased there that their idling GPU temps have dropped down into 60 to 70ºC range. But even these new reduced Dell GPU temperatures are higher than I usually see on my own MBP when its idling along. I use Marcel Bresink's "Hardware Monitor" which provides probably the most accurate readings on Macs, and also offers a "history" chart. It idles along in the 50 to 60ºC range most of the time. When genuinely pushed it will move up into the 70's, but the highest I have ever seen it was around 75ºC. This is 20 degree Celsius or better less than the Dells seem to have been running at when involved in similar activity, despite the fact that I'm also pushing an external 23" monitor as well as the internal screen.

I see a number of people over on the Lenovo boards are making much the same point. Some Thinkpads use these same cards too, but , like the MBP's, their GPU temps are generally much lower than in the Dells. Again Lenovo have made no statement about the matter and there doesn't seem to be any evidence there of higher than anticipated failure rates either.

In short, the cards are simply not being exposed to the same extreme stresses that they are in the brands of notebooks that have been identified as having problems.

Cheers

Rod

Just a correction to my previous post - the highest GPU temp I've seen on my own was 85ºC, not 75ºC.

Cheers

Rod

I have several intermittent problems on my current MBP with 8600M 256MB GPU.
- Vertical striping (usually disappears after display reset)
- Flashing when playing video
- Text not updated properly when scrolling
- Screenfull of corrupted nonsense (usually ok after reboot)

My warranty runs until March 2009 but I'm quite concerned about this problem, especially with the cost of AppleCare here in Switzerland. I'm not sure there is much point taking it in for repair to be replaced by a logic board with the same GPU chipset until Apple speak about the issue publically.

Well, Denis, you have a warranty that runs for another 8 months and NVidia have already indicated that whatever manufacturing problem existed has been dealt with. There is no real evidence of widespread problems of this kind on MBPs. I doubt that Apple will say anything unless there is. If they do, it will be to provide a solution, if it is really needed. If I was you, I'd be getting it looked at.

Cheers

Rod

Message was edited by: Rod Hagen

hello,

i think you are misunderstanding what i am saying, and again i never said that nor do i support that idea. now, i think we are drifting off topic here, so rather than bickering, why don’t we get back to helping people with the problem.

regards,

polisciguy

hello,



i am getting temps in the 90s (celsius) and i am not really pushing the machine. that is around 194 F for you americans out there.

regards,

polisciguy

hello,

just thought i would post this image that another apple forum user posted so that people can see one of the issues which may be related:

http://img297.imageshack.us/my.php?image=photonu9.jpg

another user posted a similar photo.

all the best,

polisci

That is display corruption (artifacting). Probably caused by overheating GPU, could be linked to this bad batch of nVidia GPUs.

Judging by another thread I have posted in, this problem is widespread among Santa Rosa MBPs. Most of them seem to be failing with a few months of the initial year's warranty expiring. There are correlations between heavy activity (eg 3G gaming, graphics stuff) and failure - but bear in mind, these are professional machines that should be expected to handle that stuff.

In my situation, I was three months out of warranty. I ended having to get a brand new MBP since the cost of repairs and labour was over half the new value of my original MBP. Apple's Geniuses were very sympathetic, and they waived the data transfer costs from my old HD to my new which was really sweet.

This is what I advise:

1. BACK UP EVERYTHING NOW. While my HD and the data on it were both fine, you still want to be able to access your data while your machine is in for repair, in the event that it fails.

2. BUY APPLECARE if you still can. I've bought it with my new MBP. I expect that my new MBP will fail at some point, but at least I'm covered for three years.

3. HOLD ON to your original MBP even if you buy a new one. Should there be some kind of compensation for this - which may only be in certain countries with better consumer laws - you may be able to get post-warranty MBP's repaired for free at some point.

4. DELAY PURCHASE of the current generation of MBPs unless you desperately need one, consider renting or borrowing or using another machine you own instead. If you do have to buy one, as I did, buy Applecare. You US users are lucky - you pay way less for it than those of us in most other countries.

5. ARRANGE A BACK UP MACHINE if you are currently working on a critical project for a client.

6. DON'T BLAME YOURSELF - the MBP is a professional machine that is designed to run graphics intensive applications well. You shouldn't have to feel constrained to using it as nothing more than a word processing machine/email checker. Assuming you have it in a suitable, well ventilated environment (not 100F heat or 100% humidity!), you are taking care of it and not bashing it around, and doing regular maintenance and maybe getting a yearly clean or check up then that should be enough.

Message was edited by: istara

Howdy All,

I don't agree with this "don't panic, wait and see" approach, this is just what Apple wants us to do.

Rod Hagen wrote:
NVidia, as I also indicated, have said quite specifically that the issue only affects some chips and only occurs in some system configurations:

By "system configurations" you see to be meaning (the correct use of the term) certain laptop models (i.e. systems) from certain vendors with the specified GPUs (or in some cases certain logic boards with the problematic GPUs).

+Only a very small percentage of the notebook chips that have shipped are potentially affected, and the problem depends on a combination of environmental conditions, configuration, and usage model.+ (see http://arstechnica.com/news.ars/post/20080716-nvidia-denies-rumors-of-mass-gpu-f ailures.html )

But that statement, I believe, is very different to what you said above. Personally, I think it is pure PR talk from NVidia in a time of crisis.

By "combination of environmental conditions" I think they are referring to the environment in which you use your computer rather than environmental conditions at manufacture (which the next quote discusses). I think this is just code for "if you use your computer in a hot place".

By "configuration" I think they are just referring to whether you have one of the specific models of computers that contain the suspect chips. Which is pretty obvious, you aren't going to have the problem if you are using one of the other GPUs.

By "usage models" I think they are referring to how you use the computer (GPU). I think this is just code for "if you use your computer so that the GPU works hard and gets real hot".

So in summary, if you "use your computer so that the GPU works hard and gets real hot" and/or you "use your computer in a hot place" and it contains the GPU we have mentioned then you are most likely going to have the problem.

+the failure appears related to the combination of the interaction between the chip material set and system design+ (see http://www.eetimes.com/news/semi/showArticle.jhtml?articleID=208802249 )

This is the interesting part. I thought the problem had been more specifically identified - thermal loading of one of the materials in the chip caused the problem. Anyway, if they have any sense they would have stopped using those chip materials when they identified the problem.

So I would have thought they would be able to identify the batches of the GPUs which would be susceptible to the problem and those that weren't, and would be able to make a clear statement about when people could relax and when they couldn't.

Of course, admitting this would be admitting that all those GPUs manufactured before they made the change would be susceptible. The computer manufacturer would then have an obligation to repair the machines if they failed inside or outside of warranty.

The specific "system designs" which have so far been shown to suffer as a result are made by Dell and HP. Is Apple also affected? Well, we simply have no evidence that they are.

If think the evidence is quite clear. A large number of the MacBook Pros with this specified chips have their GPUs failing (no video out) in a way that has been identified as a result of this problem (they show up as the incorrect GPU).

So, as I said, don't panic (and don't create panic in others). When similar issues HAVE affected Apple in the past, (with for example the very widespread capacitor issues that caused problems on a broad range of computers from many different manufacturers in 2004/5) Apple introduced an "extension repair" program once the nature of the problem had become clear. I see absolutely no reason to suspect that they will behave any less properly here, IF the issue turns out to be a genuine one.

Ah, yes, but without a statement from Apple in this regard there is uncertainty. I am coming to the end of my one year warranty period. Do I purchase AppleCare for a machine that may be susceptible to this problem? I don't know.

If my machine is not susceptible then the decision would be a regular Applecare or not decision. However, if it is susceptible and Apple may know this then I am funding Apple's repair program when I shouldn't have to.

Apple should have made a statement as soon as they had the facts from NVidea. If they don't have the facts already then something is also wrong with NVidea or Apple. Perhaps NVidea still don't know exactly what is wrong or perhaps NVidea and/or Apple are stalling.

I think the time for "determining whether this is a genuine problem or not" is well and truly over.

Apple need to make a statement and act now.

Cheers,
Ashley.