- May 26, 2004
- 3,751
- 8
- 81
http://www.xbitlabs.com/articl...twinturbo-musashi.html
If you skip to the important portion on page 12:
http://www.xbitlabs.com/articl...nturbo-musashi_12.html
It's clear that the stock cooler is much better at cooling the VRMs. To get the same VRDM temperature as the stock fan at 25%, the aftermarket solutions have to run high speed.
Nothing even comes close to as good as stock HSF at 35% speed on VRM temp, except, of course, stock HSF at 100% speed
So while aftermarket coolers may bring the GPU temps way down, the VRM temps stay as high or even higher than stock temps and may create their own problems. Their results confirm this
If you skip to the important portion on page 12:
http://www.xbitlabs.com/articl...nturbo-musashi_12.html
It's clear that the stock cooler is much better at cooling the VRMs. To get the same VRDM temperature as the stock fan at 25%, the aftermarket solutions have to run high speed.
Nothing even comes close to as good as stock HSF at 35% speed on VRM temp, except, of course, stock HSF at 100% speed
So while aftermarket coolers may bring the GPU temps way down, the VRM temps stay as high or even higher than stock temps and may create their own problems. Their results confirm this
This is why we only tested the coolers on a graphics cards working at its nominal frequencies and didn?t overclock it: even when we increased the GPU frequency only to 800MHz the card would hang during the test cycle, which didn?t occur with a reference cooler. Even the aluminum heatsinks on these components didn?t help improve the card?s overclocking potential, unfortunately.