Q6600@3.4ghz big diff in core temps

[alvine]

So yea guys I wanted to ask you is it normal that my q6600 core 1 and 2 always run warmer than the 3 and 4? At idle i see 40-40-36-36 at load I seen bigger differences like 52-52-48-48 so I have 4c difference between the cores. I have re-seated my aftermarket cooler twice now and I always have symptoms of 1 and 2 running warmer than 3 and 4.

 

[CompuTronix]

Since a Q6600 G0 is a pair of E6850's, the temperature differences you see is related to factory DTS calibrations and core loading, which is normal and expected.

 

[russss]

I see the same (same spec: G0 Q6600 @ 3.4GHz) - I'd assumed also because cores 1 & 2 see more load than 3 & 4 - under heavy load they all balance out temp-wise.

 

[V3NOM]

could cores 3 and 4 be programmed to lag like 200MHz or something behind whatever core 1&2 are set at? also, are cores 1 and two BELOW 3 and 4 or is it the other way round? i would guess the cores underneath the top ones would run hotter since they are not contacting the IHS directly to take away the heat.

 

[JDocs]

The 2 dies are placed side by side not on top of each other. That'd just limit overclocking like the survival of earth depended on it. This is however usual for Q6600s are they are, as earlier mentioned 2 separate chips and they may or may not have different heat tolerances.

 

[V3NOM]

err, there's not four cores on the die each contacting the IHS... look at the arctic silver application picture dude, it's basically two E6850's sandwiched one below the other.

 

[jamesgoddard]

err, there's not four cores on the die each contacting the IHS... look at the arctic silver application picture dude, it's basically two E6850's sandwiched one below the other.

Wrong

 

[JDocs]

Venom, having both on top of each other is a really bad idea. The bottom chip will heat the substrate and the top chip. The top chip will heat the bottom chip and IHS. Things would get really nasty really quickly. Also connecting the top die to the substrate while maintaining properly latency would be extremely difficult if not impossible with current technology. they have to be next to each other to be able to connect to the substrate properly.

 

[V3NOM]

yes, after looking back at the AS5 instructions, i see you are correct james.

 

[warezme]

So yea guys I wanted to ask you is it normal that my q6600 core 1 and 2 always run warmer than the 3 and 4? At idle i see 40-40-36-36 at load I seen bigger differences like 52-52-48-48 so I have 4c difference between the cores. I have re-seated my aftermarket cooler twice now and I always have symptoms of 1 and 2 running warmer than 3 and 4.

have you tried using a different temp monitor software like realtemp or even rivatuner with C2D plugin. Your temps are like to linear I think. My core temps are off as well but more like 38-40-41-45, they are never linear as sets. I have tried several temp monitoring tools and they are not the same and the cores are rarely if ever matching temps. Their temp has to do with their relative location on the die and the load of each individual core among other things.

 

[JDocs]

Warezme, mine are also linear as sets. Its quite common for Core 2 Quads to do this.

 

[rojito]

Forgetting all that tech stuff, just go with what computronix said. You could always make sure your HSF is properly attached (watch all four pins), but I can tell you that all Q6600 I've seen (about 4) including mine, have different temperatures on each pair of cores (2-5 °C)

 

[warezme]

Warezme, mine are also linear as sets. Its quite common for Core 2 Quads to do this.

hmm, maybe cause mine is watercooled and just how the water intakes hit them but I rarely have two cores match temps

 

[gamerk316]

Guys, this is normal. Every quad has a 4-5 degree diffrence on the last two cores.

Its normal, don't worry about it.

 

[V3NOM]

hmm... be good if they were all the same temps though, you could overclock those other two cores more

 

[q6600temp]

This is common, just add all of the cores temperatures and divide them by 4 and you will get an average.