New Posts  All Forums:

Posts by u3b3rg33k

If I wanted an adhesive that was thermally conductive, I'd use a thermally conductive adhesive:http://www.arcticsilver.com/ta.htmin fact, I have. I watercooled an old mac (back when they were easy to overclock), and I needed extra cooling on the memory controller, but there were no mounting holes near it. thermal adhesive solved the problem (FYI it's very permanent).
those outer two heatpipes don't look like they contribute much.
from the look of that whatever sensor chip they put on the card sounds to be dead. that, and the lack of a reported RPM probably means the card thinks the fan is running too slowly, and cranked it to 100% to protect itself. I think you're out of luck on this one.what do you see in catalyst control center?
here's some more:http://www.doitpoms.ac.uk/tlplib/optical-microscopy/preparation.phpand with a little (ok maybe a lot of) poetic license, here's the last one under a scanning electron microscope:NOT REMOTELY PERFECT!
You're missing what we're saying. we're saying that an ideal metal to metal interface will blow metal - TIM - metal out of the water. an ideal metal to metal interface doesn't exist, therefore you can't try it.
Pull up GPU-z and see what it thinks the temp and fan % are?
absolutely. they exist in multiple forms:here's a non-computer version:the general downside of passive heatsinks is that they have to be very large to get the job done.WRONG!!! just because perfect metal to metal contact cannot be achieved doesn't mean it's not way better! "in theory" with perfect metal to metal contact a pad/paste makes it WORSE!
shorting out a fan header will not result in an RPM signal being picked up by the motherboard. FYI.
the point of a card with low clocked 50+ cores that talks through the PCIe bus is likely not going to be single thread performance...
GTX 660 only needs a single six pin...
New Posts  All Forums: