Noob questions on X2 temps

[Angilion]

i) How high is too high?
ii) Is it normal for one core to run up to 7C hotter than the other?

I'm getting 56/63 at stock under full load and that's much hotter than any CPU I've had and monitored before. Winding my case intake and CPU cooler (which doubles as case extract) fans up to max only drops the temps to 56/61 - just 2C on just one core going from minimum fan speed (1100 rpm) to maxium (2000rpm).

 

[Angilion]

More info, as requested:

The board is an Asrock 939 Dual-SATA2.

Cooling is a 120mm Akasa Ultra Quiet ("Amber") as case intake and an Akasa EVO 120 with the same model fan as CPU cooler and case extract (the EVO 120 design has the fins and fan pointing at the rear case extract fan mounting, a couple of inches away from it).

Fan speeds are adjustable from 1100rpm to 2000rpm for the case fan and something similar for the CPU/extract fan. Varying from min to max has no effect on the temp of one of the cores and drops the other by 2-3C.

Core Temp 0.94 pic:

I'd been using 0.95 without any problems, but 0.94 does show more information...so what is the point of 0.95?

Speedfan 4.32 pic, to show other temps. I'm not sure which is which. Not taken at the same time, so the core temps vary slightly. The other temps are the same, though.

The cores idle at 39-41C, too. The core that reaches 62C jumps 14 or 15C as soon as load is applied and drops 14-18C as soon as it is removed, whereas the core that reaches 56C immediately jumps and drops only 8 or 9C.

I think something is awry. I don't even know if these are true readings or not.

 

[Angilion]

I mounted the HDD slightly above the case intake fan, so it's thoroughly cooled.

I'll check the heatsink temp tomorrow, as I'm about to go to bed. If the heatsink is fairly cool, I'll take the chip out again, clean everything again and do the ruler test to see if the IHS is bent. Maybe I'll try spreading the MX-1 in a thin layer, even though the instructions say not to. Maybe not - MX-1 isn't very spreadable.

What do you think that the very minor difference in temps when the fan speeds are doubled indicates? Airflow is obviously not the issue.

Is it possible that one or more of the heatpipes could have been broken (internally) when I swapped CPUs? How fragile are they?

Thanks for the straightforward answer about the different core temperatures.

What's really bugging me is that I know this chip will overlock a lot, but I don't want to overclock it at all with those temps.

 

[Angilion]

Iirc did'nt the Evo120 get rather poor reviews when used with dual cores? In any case thouse temps are very poor, the 4 pipe dual core stock AMD cooler would perform better than that Evo there (im not joking )

I don't know - a quick look found a couple of very good reviews and a lot of rubbish that had nothing to to with them. Before I buy a new cooler...do you have any links to any of these rather poor reviews?

 

[Angilion]

No, as i said iirc. Havent seen or read any review of said cooler since it hit retail; which was quite a while back now tbh. Did it have 'Dual Core supported' stamped on box or maybe instal manual, if not then...

Akasa's website says it's for dual cores. No direct link, but it's on http://www.akasa.co.uk/ as AK-920.

Why is specifically dual cores an issue, as opposed to the amount of heat generated?

 

[Angilion]

Ahhh I see. Gonna have to download that core temp program as Everset gives me one figure, Asus PC probe gives me an entirely different one.

I'm seeing a little less than 60 ish under load overclocked to 2600Mhz on Asus Probe and 65 during full load on Everest. I have no idea which is reporting an accurate figure !!

Wrongfully assumed that OP was worrying about 60 ish in an overclocked environment for which I still think that temp range is OK with that kind of overclock tbh.

25% o/c adds 3C. I think the chip has a lot more headroom, but I don't like the temps at all.

 

[Angilion]

From the looks of it the Evo 120 could very well be the culprit. I read a couple of reviews and it doesn't look very good and possibly it's not able to remove the heat produced by dual core CPUs quickly enough resulting in heat saturation especially when the CPU is stressed. Probably the heatsink's aluminum fins won't be that hot at 60c load and rather than blame it on the heatsink mounting I'd be more likely to blame it's efficiency when it comes to dissipating a lot of heat.

Apparently there was a version 1 and version 2 Evo 120 and the main difference being bigger heatpipes on the version 2 as well as improved mounting.

Buying a new heatsink might be the only way to know for sure and I'd recommend going with one of the big towers that blow out the back just like the Evo 120 so as to maintain the low temps elsewhere in your system.

I asked in the shop when I bought the X2 and was assured that the EVO 120 was more than good enough for cooling it. Coupled with the good reviews of it when it came out and my good experience with it, I considered that enough reason to stay with it.

Having spent some more time looking at reviews, I agree the EVO 120 is probably the problem. It seems that the limit to the amount of heat it can dissipate has been exceeded by my X2. Below that limit, it's an excellent cooler (it beat most other coolers when it came out) but the performance drop when that limit is reached is precipitous.

Bah. If I had know this beforehand, I wouldn't have bothered with the upgrade.

 

[Angilion]

More confusion...OcUK sell the EVO 120 and the average rating for it from customer reviews is 5 stars (the maximum).

EDIT: In the customer reviews for the X2 3800, there's one customer using an EVO-120 on an o/c X2 3800 and another using the earlier EVO-33. Both claim low temps.