sponge_bob_128
n00b
- Joined
- Sep 21, 2011
- Messages
- 21
I am kind confused by these "new" monster coolers, the core temperature utils are showing the cores overheating, yet the cooler is only luke warm (maybe 30C) (fins or top of the base). Old school pin coolers ran much hotter and running high rpm fans cooled them off. Yet from my experiment (holding a higher rpm fan pulling air through the Noctua) I don't think you will get much of a core tempurature drop.
I say that my i7 is heat limited as 4.5Ghz and 1.35V it is in the 70C's, but crank it to 4.7Ghz and 1.375V and it goes into the 80C's (Prime95-10K FFT-8 threads). I am surpised that the monster Noctua cooler is the limiter. Yet I believe the Noctua is holding the base to a low tempurature and since I believe I have applied the thermal grease reasonably, the processor metal cap/IHS should be at a fairly low temperature. This implies to me that the heat conductance from the core to the IHS is not sufficient at these high power levels (at least for my chip). In the old days, we used to be able to rip off the IHS and put the heat sink directly on the exposed core for a good drop in temperatures, but no longer. The only solution seems to be to drop the heatsink block temperature to even lower temperatures. But requiring such extremely low temperatures seems to be a flaw in the i7, i.e. the Noctua can cool higher wattage processors, yet the low power i7 2600K cores overheat.
Now it is likely a few more speed clicks up would smash into voltage or stability limits, but at least for my processor, the limit is heat.
I say that my i7 is heat limited as 4.5Ghz and 1.35V it is in the 70C's, but crank it to 4.7Ghz and 1.375V and it goes into the 80C's (Prime95-10K FFT-8 threads). I am surpised that the monster Noctua cooler is the limiter. Yet I believe the Noctua is holding the base to a low tempurature and since I believe I have applied the thermal grease reasonably, the processor metal cap/IHS should be at a fairly low temperature. This implies to me that the heat conductance from the core to the IHS is not sufficient at these high power levels (at least for my chip). In the old days, we used to be able to rip off the IHS and put the heat sink directly on the exposed core for a good drop in temperatures, but no longer. The only solution seems to be to drop the heatsink block temperature to even lower temperatures. But requiring such extremely low temperatures seems to be a flaw in the i7, i.e. the Noctua can cool higher wattage processors, yet the low power i7 2600K cores overheat.
Now it is likely a few more speed clicks up would smash into voltage or stability limits, but at least for my processor, the limit is heat.