The nvidia GPU issue thats all the rage in the news..

What exactly is the deal that came up a few months back, and again now?

ie: Its to do with heat cycling.. repeatedly heating and cooling the GPU and it goes bad.

What parts are being nailed, and how does that impact us?

ie: I've got an 8700 and 8800GTX (upgraded myself) .. are one or both cards effected by this? And assuming so, is the lifspan shorted to a few years, or are we expecting a lot of us to go kaput any time now, depending on usage patterns?

I would think with such a massive problem, there should almost be a recall and/or replacement cards, but we all know it'd never happen. Discounts on replacement cards at least, and so on. (Of course, they're still making defective cards I bet?)

What is the best option for us all?.. or ignore it and pray?

OR is it just overblown, and its not a big deal?

jeff

 

8700MGT is effected, 8800M dono....

 

The best option is to monitor your gpu temperatures and speed (and note if it every overheats or downclocks), possibly purchase an extended warranty if you're worried, and send your notebook in for repairs if it acts up. From what I've seen, all or almost all 8xxx and 9xxx chips are affected and nVidia hasn't released specific info. There's no fix for it short of nVidia recalling ~30million chips, which isn't happening, or extending warranties for the cards on their own, which could happen. If it really worries you (honestly, you're not alone, and with so many people and defects expect a class action law suit and/or eventual large-scale warranty handling by nVidia) and you have the option, switch to an ATI card.

 

I'm concerned about this too. I don't really think there's much we can do. Just wait and see. In the more recent news releases, it does seem like the 8800m gtx is at risk. I remember they said initially that the problem was contained to certain laptops which couldn't keep the GPUs cool enough and that Sager was safe due to their exceptional cooling systems. That may no longer be the case if the problem is something more fundamentally linked to the card itself, not just keeping it from overheating. Hopefully nvidia gets their act together with this whole thing

 

there is one thign to do, get the OCZ whitebook

 

The strange thing is that, at least in the Clevo forums, we barely see any defects with video cards. I mean, the D900 series are out for about 3 years now, and still we don't see a lot of failure reports.
So maybe it is true that Clevo has less risk...

 

Here's hoping, eh?

 

This is very possible, if only for the fact that the heatsinks + cooling systems are so much better in Clevos. The defect results from repeated heating/cooling cycles and is a problem with the solder used in the chips/nearby. I'm assuming that systems that run these cards at 100C+ have a far higher risk of cracking the solder, so yeah, Clevo's might be pretty immune to nVidia card failures.

Another reason to monitor your temperatures and do all you can to keep the system running cool while gaming.

 

this really confuzzles me.. isn't nvidia doing anything about it? or are they doing something but too slowly such that millions are getting affected by it as time goes by?

 

We were given advice to heat up the video cards a little bit before gaming, then don't shut off the system right away and let them cool off in a gradual state. Perhaps, for those who can, having a Vista Dream Scene on the desktop might just be that solution. It does use the graphics card to a small extent in comparison to the gaming.

 

The XR5 that I RMA'd had a fried motherboard and problem with its graphics card. PCMW wouldn't give me details on what happened to the card other than it was not outputting a correct signal, could have been a heat problem.


/Saad