Dell reveals nature of NVIDIA GPU issue
Early in July we carried news that NVIDIA would shave $200 million off of their 3Q08 profits to handle the burden of repaying OEMs for faulty mobile GPUs. At that time there was considerable speculation as to what aspect of the GPU design was responsible for the significantly increased failure rates.
Courteously, Dell has not only stepped forth to provide the world with an answer, but is also looking to stem the tide. Significantly, the issue appears to be “a weak die / packaging material set, which may fail with GPU temperature fluctuations.” This confirms previous hypotheses that blamed a faulty substrate.
In order to impede the higher than normal RMA rate, Dell has also designed and released several new BIOS versions that “modify the fan profile to help regulate GPU temperature fluctuations.” It is hoped that these modifications will reduce the amount of warping the packaging material endures amidst changes in temperature.
Great. This band-aid should keep my 1520 going until the warranty expires.
Sort of what many others are saying. Some are refusing the upgrade so that the unit will fail and take advantage of the warranty.
One even suggested a hair drier ...
To heat it or cool it? Because hair driers do have cool functionality..
Serious question?
More or less. More [serious] if the thread were centralized on DIY Faltered GPU Cooling.
Srsly tho.. I'd like to know.
Well they did mention it was temperature fluctuations that caused the failure... perhaps switching from heating to cooling on your hair dryer is the way to go.