I3 350M, large difference in the two core temperatures

[idata]

I recently purchased a Dell Inspiron laptop with I3 350M processor and felt that it was running considerably hotter than the older core2duo laptop. And the CPU fan was ON all the time. Using Core Temp utility I found that core # 1 temperature was about 17 C higher than that of core # 0 even though the load on the two cores was similar . I believe this is a problem. I need suggestion. Screen shot of Core Temp is attached.

Thanks

Subsequently checked that when the load on core # 0 was much higher it still remained cooler than core # 1. The next screen shot shows this.

Thanks again.

Message was edited by: SPSH

[idata]

That seems to much of a difference, i would advise taking the laptop back and getting a replacement. It could be simeply a misreading, as i doubt it is possible for teh cores that are so closer together to have such a difference in temp.

[idata]

This was purchased online and I have told Dell about it. Waiting for their response.

Yes, highly unlikely. But the trouble is I got on to this because I felt the laptop was getting much too hot. And it has been with me for just a week.

Thanks

[idata]

Sorry for the delay to write back in this post.

Dell eventually replaced the fan, heat pad and heat sink of the microprocessor. I didn't expect this to solve the problem and neither it did. With the present ambient temperature lower by nearly 15 degree C since the time the laptop was received, the CoreTemp readings essentially remain the same. Dell are still researching the issue.

Doesn't Intel respond here with a solution? Any suggestions will be welcome.

Thanks

[idata]

Hi again!

In terms of the CPU being too hot, that is not the case, it isn't overheating as it is in within the spec temps, the Tjunction is 90o for rPGA and 105 for BGA as you can see here: http://ark.intel.com/Product.aspx?id=43529&code=350M http://ark.intel.com/Product.aspx?id=43529&code=350M

Regarding the difference in cores, that is odd and the only thing i cna think is as i said before that it is a misreading or the DTS in the CPU is messing up. Whethert this will count as faulty and tehrefore replaceable, will depend on Dell and or whomever you bought it from. Doubt Intel will comment on this even if you contact tehir suport line as it is an OEM product, so in the end it is Dells responsability.

[idata]

Hello Cpt.Dogfruit,

Thanks for that prompt reply. Tjmax in this case is 90 C. So overheating was never the question. The difference in temps is.

Ok, so I may not get much help but no harm sharing my observations and views here.

1. I put a clinical thermometer right at the fan exhaust outlet and I had to remove it when the temp reached 42 C - didnt want to risk breaking it. CoreTemp reading at that time was approximately 35 and 50 C for core 0 and core 1, respectively. And the exhaust reading was nearly average of the two. Course method. But I dont expect a misreading by CoreTemp. And in any case other utilities too give similar readings.

2. Although it cant be said to be overheating, it can cause performance degradation surely and other component failure potentially. Which of the two cores temperature is taken as reference to control the CPU fan speed, CPU throttling, and system shut down? I do not know. You take the higher of the two and degrade the performance, using the lower temp endangers the CPU. And CoreTemp only looks at temps at 1 second intervals. Instantaneous temp reading of core 1 could be even higher. Complications are fairly obvious.

That this difference in core temps under similar conditions is abnormal, is clear to me from your reply and hopefully Dell will provide a solution - eventually.

As far as I can see it is either the CPU or the motherboard. I will post here as and when some new development takes place.

Thank you.

[idata]

Hi,

I mentioned teh max temp coz of tehis "But the trouble is I got on to this because I felt the laptop was getting much too hot"

In response to this: "Although it cant be said to be overheating, it can cause performance degradation surely and other component failure potentially. Which of the two cores temperature is taken as reference to control the CPU fan speed, CPU throttling, and system shut down? I do not know. You take the higher of the two and degrade the performance, using the lower temp endangers the CPU. And CoreTemp only looks at temps at 1 second intervals. Instantaneous temp reading of core 1 could be even higher. Complications are fairly obvious." Well it won't in theory cause any performance dergradation, as i the throttling, temp control all that would be taken from the DTS (which isnt teh core temps). Therefore it should not damage your CPU. But yes personally i would see it as an error.

[idata]

Hi,

DTS merely converts the analog signal from the thermal sensor located within the CPU chips into digital form and stores it in a register for any process to use it for whatever purpose - including the system control. External utilitities like CoreTemp too read the temperature from there.

I believe this is not only an error but a fault somewhere.

Thanks.

[idata]

Well, after first trying with replacing CPU fan, heat sink and the thermal pad - which didn't help resolve the issue - Dell replaced the processor as well as the motherboard. The core temps are within 3 to 5 deg C of each other.