topic Re: T9400 - Temperature problem in Mobile and Desktop Processors

T9400 - Temperature problem

idata — Wed, 03 Nov 2010 16:38:14 GMT

Hi,

I have a notebook with this CPU, and I get temperatures on Idle of about 65ºC (149ºF) and 90ºC (194ºF) on load.

The notebook manufacture say this temperatures are normal for that notebook model.

I know that the TJunction temperature for this CPU is 105ºC, but my question is: Are this temperatures safe for the CPU?

Thanks

Re: T9400 - Temperature problem

idata — Wed, 03 Nov 2010 17:15:18 GMT

Hello; The maximum temperature for your processor is 105 °C, the current temperatures of your processor are normal. In the case that the processor exceeds the 105°C of temperature it will shut down the system automatically to prevent any damage. You can check your processor specifications here: http://ark.intel.com/Product.aspx?id=35562 http://ark.intel.com/Product.aspx?id=35562

Re: T9400 - Temperature problem

idata — Thu, 04 Nov 2010 11:50:22 GMT

You can try to set the fan to a higher speed so that the temps can be lower. The high temps shouldn't be a real problem, other than the longevity of the CPU isn't motivated by the high temps.

You can use http://www.almico.com/sfdownload.php Speedfan or similar to access the internal sensors and adjust the fan speeds (if compatible).

Re: T9400 - Temperature problem

idata — Mon, 15 Nov 2010 15:21:33 GMT

> Hello; The maximum temperature for your processor is 105 °C, the current temperatures of your processor are normal. In the case that the processor

> exceeds the 105°C of temperature it will shut down the system automatically to prevent any damage. You can check your processor specifications here:

http://ark.intel.com/Product.aspx?id=35562 > http://ark.intel.com/Product.aspx?id=35562

Hello Sergio,

Tthis relates to my recent question.

Does this refer to like "instant damange", like the cpu doesn't instantly pick up damange up to that temp?

But what about accumulated tear and decereased lifetime?

Consider a cpu constantly operting at ~ 95C or so, although it works, shouldn't we epxect a drastically shorter lifetime or probability of failure of such design than one hat is cooled to say max 65 at max load?

I've searched the net and rarely does anyone discuss this. I'm not interested if the cpu drops dead if I hit 90C, I want to know if I will reduce the lifetime by doing it - if so - I still would not consider it good practice to run it so hot?

Any more numbers on this would be greatly apprecaited. I'm having a discussing with a designer that seems to suggest that anything below the MAX value is "good" cooling. Now that doesn't make sense to me, and I'd like official intel comments.

/Fredrik

Re: T9400 - Temperature problem

DSilv11 — Tue, 16 Nov 2010 02:01:24 GMT

All the newer generation processor include thermal throttling.

This mean you can't get the CPU too hot. (blow torches excluded!)

Under anything near a normal operational condition, if the CPU \ cooling solution starts getting too hot, the processor will throttle slowing itself down to a cooler operational temp. (you also loose horse power which makes this a bad thing to rely on.) If you manage to get the CPU hot enough, the CPU will assert PROCHOT and shut down completely.

As to the impact of extended operation at higher temp.

This is a discussion that frequently comes up between the System engineers designing the chassis and cooling and the board hardware engineers.

Heat kills and the less heat in any electronic component the better.

The issue I find more often is the other components on the board rather than the CPU / chipsets.

As you note above, The maximum temperature for your processor is 105 °C, BUT, very likly the maximum temp of the capacator sitting above, below or next to the processor is 85 deg C. I susp[ect that extend operation at high temp will likly cause the supporting components to fail lon before the processor starts showing issues.

Re: T9400 - Temperature problem

idata — Tue, 16 Nov 2010 07:27:44 GMT

Thanks for both your responses, it may not have answered the question but thanks for your pointers: I'll try to contact intel. I'm quite sure I skimmed the thermal desgin papers for one of the cpus before and it just contained qualitative information but not the quantification of lifetimes needed for a specific tradeoff choice.

> This is a discussion that frequently comes up between the System engineers designing the chassis and cooling and the board hardware engineers.

Yes, this is the right context of my question. I am discussing with someone who has "solved" the noise problem, but reducing fan speed, and then about temp the argument they use is the "look it's below 105 - which is intels spec - so it's fine". Now that seems like a lame argument ot me and I would like to quantify the impact of the temp increase if possible.

I have some experience in the past where customers have made non-optimal but still obviously "working" mounts that hinder airflow to cpu (but didn't stop it completely of course), and we've seen an increase failure rate of cpu's popping after only 2-3 years on a few units (although that was older type cpus). So I think this is a relevant question to ask. I also understand that due to the long timetime of a cpu it's hard ot make real tests, but probably one can make tests at extreme temperature and maybe extrapolate the lifetime by theory to the normal ranges.

/Fredrik