Nel said:
I'm certain that my CPU (P4 3.06gHz) is running far too hot and is the cause
of my unstable PC. The manufacturer claims the temp is normal for this type
of system - which is more like a laptop than a desktop, as the motherboard
and cards are built in behind the TFT screen and the hardrive is built into
it's base.
They actually say 108-113C is 'normal', or do they say that it's normal for
their system to report funky temps?
Well, first, is it a desktop P4 3.06 gig or a mobile and do you know what
the temperature reading is coming from? I mean, case temp, as would be the
proper measurement for a desktop P4 or on-die temperature diode, as would
be needed on a mobile?
The max case temp for a desktop P4 3.06 gig is 69C, so your temps are
obviously not 'normal' for that measurement, regardless of what kind of
cute case it's inside of.
If it's a mobile, then the on-die thermal diode is monitored and that is
inherently a higher reading than a 'case' temperature for the 'same'
condition because the die is hotter than the case (internal thermal
resistance from the die to case). If you're measuring DIE temperature then
the maximum operating spec is 100C and above that the P4 should throttle
it's speed back in an attempt to lower it's consumed power, hence lowering
the temperature, with 125C being shutdown (average numbers as I didn't look
up a 'P4 mobile 3.06 gig' specifically because there isn't enough variation
between the versions to matter substantially).
Even here, your number of 108C to 113C is above anything that could be
considered 'normal'. If those numbers are real, it would suggest the DIE
temp is what's being measured and that your processor is 'normally' in the
over-temp throttle range (depending on how accurate the temp reading really
is) and on the edge of thermal safety shutdown; which WOULD explain your
observed symptoms.
They are correct that mobiles (which I presume they're claiming their
system is 'like') operate the CPU at higher temperatures (albeit not as
much as one might think due to the spec being a DIE temp rather than CASE
temp) than the typical desktop and it's also true, even though not talked
about much, that the thermal design specs for systems of that 'type' are
based on 'real applications' (Intel's terminology) and not continuous 100%
loads (continuous 100% load may overload the heatsink and put it into
throttle down), but there is no way, for even that "type of system," that
it should have a processor temperature over 100C under 'normal' operation
in a room temperature environment unless it's, put bluntly, improperly
designed, or you've got it somehow placed in an environment above it's
thermal specifications (like say if the SYSTEM spec is up to a room
temperature of 30C and you've got it in a 40C environment, or you've
blocked the cooling vents with something, or some other 'not normal'
condition).
And, under 'normal' conditions, it should certainly NEVER hit thermal
shutdown (which I suspect is what causes your system 'instability'). Nor
should it hit thermal shutdown even if you put a permanent 100% load on it
either as the thermal solutions should be able to handle the heat from it's
throttle down mode (with 100% load). It should simply operate at a lower
effective speed.
To quote the Intel spec sheet for 'mobile' (as in a notebook) processors.
"... Analysis indicates that real applications are unlikely to cause the
processor to consume the theoretical maximum power dissipation for
sustained time periods. Intel recommends that complete thermal solution
designs target the Thermal Design Power (TDP) indicated in Table 23. The
Intel Thermal Monitor feature is designed to help protect the processor in
the unlikely event that an application exceeds the TDP recommendation for a
sustained period of time....
With a properly designed and characterized thermal solution, it is
anticipated that the TCC would only be activated for very short periods of
time when running the most power intensive applications. The processor
performance impact due to these brief periods of TCC activation is expected
to be so minor that it would not be detectable. An under-designed thermal
solution that is not able to prevent excessive activation of the TCC in the
anticipated ambient environment may cause a noticeable performance loss,
and may affect the long-term reliability of the processor. In addition, a
thermal solution that is significantly under designed may not be capable of
cooling the processor even when the TCC is active continuously."
The last description of a "significantly under designed" thermal solution
sounds like what you've got. If your numbers are real, it looks as if it is
unable to properly cool the processor under even normal conditions so that
if a significant load occurs it's swamped even more and, sometimes, to the
point of the processor's protective thermal shutdown.
According to Sisoft Sandra, the CPU runs at 108-113'C during normal
conditions an the top of the unit is hot to the touch. Can I rely on this
software's reading?
What do you mean by 'normal conditions'?
I've also heard that a P4 powers down when it reaches are certain temp to
prevent burn out - is there any software that can prove that this is
happening?
The P4 should throttle when the DIE temp, not case temp, reaches 100C and
shutdown at 125C.
