difference in core temperature is 6c. Should i worry?

I have a Lenovo G530 laptop. I'm about to give it to my brother. I bought it with a T4200 CPU. I downgraded it to a T7300, and today I upgrade it to a T4400. According to cpubenchmark.net the T4400 is much faster than the T7300.

I know that I applied the thermal past properly. Also the size of the die on the T4400 is a bout 2/3 the size of the die on the T9900. So my guess is that my T4400 stated it's live as a Core 2 duo with 3 or 4MB L2 cache, but was sold as a Pentium dual core with 1MB L2 cache because the rest of the cache is defective.

I run orthos for 15 minutes. The max temperature is 56c in one core and 62c in the other one. This is by the way cooler than the T7300 which was undervolted.

Is it fine to have this difference knowing that there's nothing I can do to reduce it?

 

Try re-pasting it. I had the same issue with my T9900 and re-pasting solved it. The slightest uneven distribution of the paste or an air bubble can cause that difference. Nothing to worry about though.


--

 

It's perfectly normal.

 

I re-applied the thermal paste 3 times and the result is the same. There is no difference in temps with the T4200 nor the T7300, only with the T4400. So I'm confident that I applied the paste properly

If it is not dangerous than I'm fine with the difference... thanks

 

The sensors aren't that accurate, they don't need to be. It's quite normal for one to be slightly different to the other, though 6 degrees is quite a bit.

 

Perfectly normal. On my i7-2720QM quad core I have one that is consistently about 6-8C higher than the others.

 

The first core should be more heavily loaded than other cores in normal circumstances. I'm not sure if 6 degrees is "normal" or not, though. Are you sure your heatsink is making even contact? Since you've repasted it 3 times you should have had a good visual indication when you took the heatsink off.

 

Windows shuffles things around the cores periodically to address that specific problem and to prevent uneven core use which could lead to an early death.

 

The first part is true, the second is not. A core is not going to die of heat death or exhaustion. It will throttle before it could damage itself.

However in newer intel core architecture, at least from what I've seen, core 0 can often be the hotter of the cores and the most loaded. It definitely varies, and the load does get spread across the cores for multithreaded stuff.

Again I'm going to point at the heatsink as the main suspect.

 

Yep, an uneven heatsink is the most probable culprit. The temps on my 920xm vary from 1-2C when idle and same on the p8600 in my N50. I get more significant temps on one core if i run a single or dual threaded benchmark on the 920xm though.

 

Intel's core temperature sensors were only designed for reasonably accurate thermal throttle and thermal shutdown control. They were never designed for 100% accurate core temperature reporting so don't lose any sleep when your two core temps don't line up. 45nm Core 2 sensors are the least accurate of all the sensors that Intel has used. A difference of 6C is not unusual and doesn't mean anything.

 

That's nice to know

OK then. Now, I need to figure out if the T7500 is faster than the T4400. I'm asking the question in another thread. This laptop will be handed over to my brother in a week and I want him to have the best of these two CPUs. I happen to own both of them so there is no cost involved.
The thread is here: T7500 vs T4400

 

They should perform very closely to each other. The margin is so small it should be hardly noticeable.

Pure CPU and Memory Test results:
T7500
Hewlett-Packard HP Pavilion dv9700 Notebook PC : Geekbench Result Browser

T4400:
Hewlett-Packard HP Pavilion dv6 Notebook PC : Geekbench Result Browser

 

thanks for the links