Processor temperature problem

S

SANTANDER

My MB is Socket 478 for Intel P4 processor, Intel 845GV chipset;
Intel Celeron 2.66GHz CPU ( Processor core: Prescott-256 )

The CPU temperature constantly stay near 57 - 59œC that seems abnormally to
me. I just asked some tech specs, and they told, that this CPU model usually
is heated and noisy (as fan constantly tries to cool the CPU). I am not sure
is this correct.

As I know from CPU specs, the max operating temperature specified is 67œC.
http://www.cpu-world.com/CPUs/Celer...67256 - NE80546RE067256 (BX80546RE2667C).html

Is there a way to solve this, or it's normal?

Thanks.
 
P

Paul

SANTANDER said:
My MB is Socket 478 for Intel P4 processor, Intel 845GV chipset;
Intel Celeron 2.66GHz CPU ( Processor core: Prescott-256 )

The CPU temperature constantly stay near 57 - 59œC that seems abnormally
to me. I just asked some tech specs, and they told, that this CPU model
usually is heated and noisy (as fan constantly tries to cool the CPU). I
am not sure is this correct.

As I know from CPU specs, the max operating temperature specified is 67œC.
http://www.cpu-world.com/CPUs/Celer...67256 - NE80546RE067256 (BX80546RE2667C).html


Is there a way to solve this, or it's normal?

Thanks.

Cooling can be a complicated subject, especially if you're
trying to fix a pre-built computer (Dell/HP/Acer etc). Some
of those, leave no room for improvements, due to the usage
of a single fan to cool everything.

Your processor might use 73W under some conditions. That would
be when the processor is running at 100% utilization. When
sitting idle, the 90nm generation of processors had high static
leakage, so the processor might still run hot, even when it
isn't doing anything.

http://processorfinder.intel.com/details.aspx?sSpec=SL7TH

If the OS is using the HLT instruction in the idle loop, that
helps reduce the operating temperature. It does that normally,
and what I discovered, is that my antivirus software overrides
that feature, and causes more power to be wasted than is necessary.
Using RMClock from Rightmark and enabling the HLT option in
there, reduces the CPU temperature when the computer is idle.
(In my case, the CPU goes from being 12C hotter than the case
air temperature, to only 1C hotter than the case air temperature.)
But using HLT in the idle loop, does not improve the heating
caused by running the CPU at 100% loading. The temperature
elevation there, remains the same, because HLT only improves
the performance of the idle loop.

One solution, is to use a different CPU cooler. But some of
these, cost as much as a Celeron. This one weighs 800 grams without
the fan installed, and comes with a backing plate. You have to pull
the motherboard out of the computer case, to install the backing plate.
The heatsink is so tall, that some computer cases aren't wide enough
to accept the cooler. So for some people, a new computer case is
required. (Read the customer reviews, for more information.)
This is an example of a *big* cooler.

Tuniq Tower 120 P4 & K8 CPU Cooler $50
http://www.newegg.com/Product/Product.aspx?Item=N82E16835154001

Another fix you can try, is to remove the current heatsink, from
the processor. Clean both metal surfaces with isopropyl alcohol and
cleaning clothes, being careful not to get residue into the CPU
socket. Then apply fresh thermal paste. An example of a thermal
paste product is here.

Arctic Silver Ceramique $5
http://www.newegg.com/Product/Product.aspx?Item=N82E16835100009

Installation
http://www.arcticsilver.com/pdf/appinstruct/cmq/ins_cmq_singlecore_wcap.pdf

Re-applying the thermal interface material, between the CPU and
heatsink, may slightly improve cooling performance. Also,
if there are significant amounts of dust clogging the cooling
fins, then removing that dust can also help the cooling
performance.

Finally, whether this high operating temperature is making a difference,
can be determined with RMClock as well. In this article, they
experiment by turning off the CPU fan, to demonstrate the response
of the CPU.

http://www.digit-life.com/articles2/cpu/intel-thermal-features-p4.html

In this picture, you can see the declining line, when the CPU temperature
is higher than the internal "throttling" temperature limit. The jagged
purple line in the upper trace, is the effective CPU operating speed,
and the user of this computer would find it running slower, when the
CPU overheats. The purple line in the display should remain flat and
aligned with the other line in the display - that indicates all of
the CPU performance is currently available.

http://www.ixbt.com/cpu/intel-thermal-features/prescott_3b.png

RMClock can be found here.

http://cpu.rightmark.org/download.shtml

Paul
 
G

Gurney

My MB is Socket 478 for Intel P4 processor, Intel 845GV chipset;
Intel Celeron 2.66GHz CPU ( Processor core: Prescott-256 )

The CPU temperature constantly stay near 57 - 59œC that seems abnormally to
me. I just asked some tech specs, and they told, that this CPU model usually
is heated and noisy (as fan constantly tries to cool the CPU). I am not sure
is this correct.

As I know from CPU specs, the max operating temperature specified is 67œC.
http://www.cpu-world.com/CPUs/Celer...67256 - NE80546RE067256 (BX80546RE2667C).html

Is there a way to solve this, or it's normal?

Thanks.

Why ask HERE? Your problem is not related to the OS and is purely
hardware related.

Ask elsewhere
 
L

Lil' Dave

SANTANDER said:
My MB is Socket 478 for Intel P4 processor, Intel 845GV chipset;
Intel Celeron 2.66GHz CPU ( Processor core: Prescott-256 )

The CPU temperature constantly stay near 57 - 59œC that seems abnormally
to me. I just asked some tech specs, and they told, that this CPU model
usually is heated and noisy (as fan constantly tries to cool the CPU). I
am not sure is this correct.

As I know from CPU specs, the max operating temperature specified is 67œC.
http://www.cpu-world.com/CPUs/Celer...67256 - NE80546RE067256 (BX80546RE2667C).html

Is there a way to solve this, or it's normal?

Thanks.

Aside what Paul provided, I never assume anything. What's the room temp and
mobo temp?

Celerons generally run cooler than their Pentium cousins. A PC that I own
has a Intel 845 PE chipset with a P4 @ 2.4 GHz. CPU temp is 47 C. Yes,
its a Prescott.
 
M

M.I.5¾

Gurney said:
Why ask HERE? Your problem is not related to the OS and is purely
hardware related.

Ask elsewhere

Gurney is our resident ****wit who just cannot grasp the simple premise that
this is a good a place for you hardware question as any.

Your question is welcome.
 
S

SANTANDER

Lil' Dave said:
Aside what Paul provided, I never assume anything. What's the room temp
and mobo temp?

Celerons generally run cooler than their Pentium cousins. A PC that I own
has a Intel 845 PE chipset with a P4 @ 2.4 GHz. CPU temp is 47 C. Yes,
its a Prescott.
---------

CPU Temperature: 56-58 C
MB Temp: 35 - 38C
CPU FAN Speed: 3777 - 4473 RPM

Where is the problem and how to solve this? I dont want install special
large CPU cooler. I that case it's even better to replace Celeron CPU to
Pentium CPU

Thanks
 
P

Paul

SANTANDER said:
---------

CPU Temperature: 56-58 C
MB Temp: 35 - 38C
CPU FAN Speed: 3777 - 4473 RPM

Where is the problem and how to solve this? I dont want install special
large CPU cooler. I that case it's even better to replace Celeron CPU to
Pentium CPU

Thanks

Yes, but is this with the CPU running a program and using 100% of
the CPU ? Or sitting idle in the desktop, with no heavy use of
the CPU at all ?

A retail cooler might have a theta_R (thermal resistance) of
0.33-0.35 or so. The cooler I have on my processor right now,
is around 0.20, which is a bit better. The units are degrees C
per watt of power dissipation.

If the CPU was running at 100% load, and as a result, using
the TDP value of power (73W), then the temperature rise is
0.33C/W * 73W = 24C. That is the temperature above the
internal case air temperature. Which in this case, might be 35C
to 38C. That makes the final computed CPU temp as 24C + 35C = 59C,
which is roughly in the right ballpark. *But*, for a 100% loaded
CPU, not an idle CPU.

If these conditions are while the CPU is idle, then the CPU
is either burning more power than specified, the heatsink
is not working properly (making poor contact with the CPU,
no thermal paste present), or the computer case air temperature
is even hotter than the stated value. There are many possibilities,
and at this distance, I cannot see which might be reasonable.

I've read a number of posts like this, where the CPU is too
hot, and the poster is not able to resolve the reason why.
In some cases, people have a hot enough CPU, that it is
operating in the throttling territory, which causes a loss
of computing performance.

Paul
 
S

SANTANDER

Paul said:
Yes, but is this with the CPU running a program and using 100% of
the CPU ? Or sitting idle in the desktop, with no heavy use of
the CPU at all ?

A retail cooler might have a theta_R (thermal resistance) of
0.33-0.35 or so. The cooler I have on my processor right now,
is around 0.20, which is a bit better. The units are degrees C
per watt of power dissipation.

If the CPU was running at 100% load, and as a result, using
the TDP value of power (73W), then the temperature rise is
0.33C/W * 73W = 24C. That is the temperature above the
internal case air temperature. Which in this case, might be 35C
to 38C. That makes the final computed CPU temp as 24C + 35C = 59C,
which is roughly in the right ballpark. *But*, for a 100% loaded
CPU, not an idle CPU.

If these conditions are while the CPU is idle, then the CPU
is either burning more power than specified, the heatsink
is not working properly (making poor contact with the CPU,
no thermal paste present), or the computer case air temperature
is even hotter than the stated value. There are many possibilities,
and at this distance, I cannot see which might be reasonable.

I've read a number of posts like this, where the CPU is too
hot, and the poster is not able to resolve the reason why.
In some cases, people have a hot enough CPU, that it is
operating in the throttling territory, which causes a loss
of computing performance.

Paul
-------------

No, these conditions are while the CPU is idle, not at 100% load, no heavy
use of the CPU at all. Possible thermal paste is old?
But, the other strange thing, when the computer is idle, only Firefox is
opened, and after certain time screensaver appears, then CPU start running
at 100% load, screensaver freeze, computer case is heated, and fan starts
works noisy. Screensaver is default windows. Task Manager show the Firefox
use100% CPU.

Santander
 
P

Paul

SANTANDER said:
No, these conditions are while the CPU is idle, not at 100% load, no
heavy use of the CPU at all. Possible thermal paste is old?
But, the other strange thing, when the computer is idle, only Firefox is
opened, and after certain time screensaver appears, then CPU start
running at 100% load, screensaver freeze, computer case is heated, and
fan starts works noisy. Screensaver is default windows. Task Manager
show the Firefox use100% CPU.

Santander

Could you reinstall Firefox ?

Are you using the latest stable release of Firefox (2.0.0.14) ?
Be aware there are exploits for some of the older versions.

Do you have antivirus software loaded ?

Paul
 
S

SANTANDER

Paul said:
Could you reinstall Firefox ?

Are you using the latest stable release of Firefox (2.0.0.14) ?
Be aware there are exploits for some of the older versions.

Do you have antivirus software loaded ?

Paul
------------

Yes, I have lastest Firefox 2.0.0.14, antivirus is Ok, daily updated
automatically. I dont think problem are caused by virus.

Santander
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top