CNET FEATURED our Charlie's latest piece covering Nvidia's ongoing GPU engineering problems in a video report this week.
Cnet's coverage is somewhat harsher on Nvidia than our own reporting has been, but it seems to have understood what Charlie's been explaining and drew its own conclusions.
Charlie's most recent article about Nvidia chips in Apple's Macbook Pro notebooks can be read here and Cnet's video The BUZZ Report that features his story is available here. µ
You guys are going to end up in big trouble if you're not careful!
Obviously, this woman doesn't have a clue about what she's talking about, so I don't think you can use it to somehow validate your story.
As you, Charlie and everyone else knows by now, it's the combination of Bump & Underfill that is important, not just the Bump.
See this AMD Presentation - it seems pretty fair:
http://www.anandtech.com/GalleryImage.aspx?id=4788
So, to be scientific, Charlie needs to confirm what the underfill type being used in the 9600 is
- if it's High Tg, then nVidia are probably off the hook, having done the right thing to fix a bad situation
- if it's Low Tg, then nVidia are almost definitely not off the hook
- but it would need an actual Material Set expert to verify that
- if that were the case, you should probably seek a second or third opinion to confirm that nVidia are doing the wrong thing.
- for example, you could ask someone for AMD to comment on the particular Material Set being used.....
Because asking AMD would be a great idea. After all they would not have another agenda! Seriously though, maybe people will FINALLY pay attention to Charlie. I been quoting the inq's articles as amazing reasons not to pick up the OLD XPS like many Dell buyers STILL do. They buy the XPS under the assumption that it must be the best gaming laptop because of the brand and don't even realize they are still selling defective 8 series chips.
posted by : Phil, 13 December 2008
Phil, the whole point of this flew over your head. IT DOES NOT MATTER IF THEY COMPENSATED FOR THE PROBLEM. It does matter however, that they STATED to the public the problem chips don't exsist in the laptop when in fact they do which mean they are covering it up by compensating for it. It's still a defective product and people should be complaining to the CPA also and investors should be suing for a ton more now. All bets are off they royally screwed the pooch here.
By releasing that information to the public saying its not in Mac's is an outright lie. Which means the 200 million they wrote off is no longer correct. This is a major SECC issue and CPA problem now. It makes no difference if they compensated. Financial markets and the laws that govern them is the key here. And technically Nvidia is going to get in DEEP DOO DOO amd tje government should be investigating them at this point. I'd be willing to bet wiht how dishonest this has been they have cooked books to make warranty repairs not look like alot. I mean really are they trust worthy now?
There's no getting off the hook here. The chip is in the laptop. They told the stock holder and public it isn't. The SECC will most likely need to investigate Nvidia for false and misleading statements and consumers should be filing with the CPA. When it hurts the stock investor and consumers who may own them they deserve to be sued.
A friend of mine did have a laptop. Come to find out it was an affected one. Too bad for him its 2 years old. And ya think I'm buying Nvidia again? Nope this is my last card.
Terck: I think you might have missed the point somewhat - as the linked AMD presentation explains quite well, I think - it's the combination of High Lead Bumps, and low Tg underfill that causes the problem
- this was the combination in the 8 Series Chips that caused the initial problem
With current 9600 chip in the Macbook Pro, Charlie has shown that nVidia is still using High Lead. But he hasn't told us what the Underfill is
- if it's High Tg, then (according the the AMD presentation) the chips should be ok.
So, High Lead bumps aren't a problem by themselves - read the AMD presentation if you don't believe me...
Engineering is usually a compromise of different variables, and the High Lead/High Tg combination appears to be an acceptable compromise (again according to AMD)
Here are a couple relevant quotes from AMD (the self proclaimed experts on Material Sets):
"If you design for High Current, High-Lead bumps, you may need a high Tg underfill to manage the stress placed on the bumps"
"For maximum GPU stability and reliability, you must combine the right bump and underfill types. The wrong combination can mean disaster for the end quality of a graphics product"
nVidia have certainly screwed up on the second one, but Charlie won't tell us if they've still using the wrong underfill.......