Not a good source.
"Results based on readings from a 30x30mm copper interface die with 150W and 85W"
1) At extreme voltage, CPUs draw more than 150W.
2) CPU heatspreader may be close to 30x30mm, but it's the die underneath and node process that can affect temperatures even more. If transistors are very tightly packed and the die is smaller, then its harder for heat to dissipate. Case and point IVB runs way hotter than SB. Frostytech's tests wouldn't show that.
3) Position of the die under the heatspreader. Some coolers have 5-6 heatpipes which often don't help with smaller CPUs such as LGA1155 but are much more effective with larger die CPUs such as the LGA2011. Again, this wouldn't show up in their testing.
Overall, it's always better to test CPU coolers on the actual CPU, rather than perform a synthetic heat test on copper.
Xbitlabs does much better CPU cooler reviews in this regard.
You can even tell Frostytech's results are off since NH-D14 is far superior to the C-14 cooler.
2
0
[
Posted by: BestJinjo

|
Date: 04/14/12 05:49:34 PM]