Sign in with
Sign up | Sign in
Your question

Maybe I'm green but it seems to freeze quickly?!

Last response: in Overclocking
Share
Anonymous
a b K Overclocking
October 4, 2004 8:34:52 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Howdie, I'm new.
I put together a computer with a P4 2.8 It runs normally at about 40C,
but as soon as it gets pushed, the temp rises quickly and the system
freezes at about 47-50C. I don't run games or anything hardcore, I
first realised why it was happening when I watched the anti-virus
program do a full system scan. But this is hardly gruntwork, the CPU
runs at about 50% for 5 minutes.

Now I'm willing to accept the temp monitors (PC Alert 4) might not be
accurate, maybe it's running hotter because of my amateur
box-building. So from reading other posts I reckon I should buy some
gluey stuff for the heat sync and maybe put a proper fan in and all
that newbie stuff - OK I can give that a go.

My real qualm is, this is a Pentium, eh? Isn't it meant to pull back
it's workrate to avoid over-heating. While I'm here asking
overclockers (who better to ask), I'm not actually overclocking
myself. So is there something I've missed? Have I got a dud?

And is it damaging the cpu? Or is the cpu shutting itself down before
the damage?

Would love your say, I say, .s
(MSI 865GM2 mobo, nvidia geforce2)

More about : green freeze quickly

Anonymous
a b K Overclocking
October 4, 2004 3:15:37 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Simon Hobbs wrote:

> Howdie, I'm new.
> I put together a computer with a P4 2.8 It runs normally at about 40C,
> but as soon as it gets pushed, the temp rises quickly and the system
> freezes at about 47-50C. I don't run games or anything hardcore, I
> first realised why it was happening when I watched the anti-virus
> program do a full system scan. But this is hardly gruntwork, the CPU
> runs at about 50% for 5 minutes.
>
> Now I'm willing to accept the temp monitors (PC Alert 4) might not be
> accurate, maybe it's running hotter because of my amateur
> box-building. So from reading other posts I reckon I should buy some
> gluey stuff for the heat sync and maybe put a proper fan in and all
> that newbie stuff - OK I can give that a go.
>
> My real qualm is, this is a Pentium, eh? Isn't it meant to pull back
> it's workrate to avoid over-heating.

It will, if it CAN. By that I mean, if there is something seriously wrong
with the heatsink, and your comments about "gluey stuff" and "proper fan"
make me wonder, then it might not be able to keep it cool regardless of how
'slow' it tries to clock it and the last resort is, of course, 'halt'.

At first glance the temperatures wouldn't seem to suggest that but then I
don't know how it's arriving at the measurement. If it's an in socket
thermistor, with a incorrectly applied heatsink, then the die temp could
skyrocket without the thermistor sensing it.

> While I'm here asking
> overclockers (who better to ask), I'm not actually overclocking
> myself. So is there something I've missed? Have I got a dud?
>
> And is it damaging the cpu? Or is the cpu shutting itself down before
> the damage?
>
> Would love your say, I say, .s
> (MSI 865GM2 mobo, nvidia geforce2)
Anonymous
a b K Overclocking
October 5, 2004 11:20:52 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Simon Hobbs wrote:
> Howdie, I'm new.
> I put together a computer with a P4 2.8 It runs normally at about 40C,
> but as soon as it gets pushed, the temp rises quickly and the system
> freezes at about 47-50C.

How quickly is "quickly"? If it's reading the CPU die temperature, the
majority of the change should occur withing the first couple seconds after
the load is applied. If this does not occur, then you're reading the socket
temperature, which is more or less worthless in terms of determining actual
CPU temperature.

[...]
> My real qualm is, this is a Pentium, eh? Isn't it meant to pull back
> it's workrate to avoid over-heating.

The P4 should throttle to stop itself becoming *damaged* through
overheating. However, intability usually occurs before damage occurs, so
it's quite possibly getting unstable (and thus crashing) before it reaches
the point where it starts to throttle.

[...]

--
Michael Brown
www.emboss.co.nz : OOS/RSI software and more :) 
Add michael@ to emboss.co.nz - My inbox is always open
Anonymous
a b K Overclocking
October 5, 2004 11:20:53 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Michael Brown wrote:

> Simon Hobbs wrote:
>
>>Howdie, I'm new.
>>I put together a computer with a P4 2.8 It runs normally at about 40C,
>>but as soon as it gets pushed, the temp rises quickly and the system
>>freezes at about 47-50C.
>
>
> How quickly is "quickly"? If it's reading the CPU die temperature, the
> majority of the change should occur withing the first couple seconds after
> the load is applied. If this does not occur, then you're reading the socket
> temperature, which is more or less worthless in terms of determining actual
> CPU temperature.
>
> [...]
>
>>My real qualm is, this is a Pentium, eh? Isn't it meant to pull back
>>it's workrate to avoid over-heating.
>
>
> The P4 should throttle to stop itself becoming *damaged* through
> overheating.

true.

> However, intability usually occurs before damage occurs, so
> it's quite possibly getting unstable (and thus crashing) before it reaches
> the point where it starts to throttle.

Well, it's supposed to throttle before it becomes unstable, it should just
be 'slower', but that depends on the thermal solution being good enough to
cool it when throttled.

>
> [...]
>
> --
> Michael Brown
> www.emboss.co.nz : OOS/RSI software and more :) 
> Add michael@ to emboss.co.nz - My inbox is always open
>
>
Anonymous
a b K Overclocking
October 5, 2004 11:20:54 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

> Well, it's supposed to throttle before it becomes unstable, it should
> just be 'slower', but that depends on the thermal solution being good
> enough to cool it when throttled.

Actually, you don't need a cooling device to keep it stable when
throttled down (at least for a few seconds) - have a look at this Tom's
Hardware article and its associated video:

http://www20.tomshardware.com/cpu/20010917/index.html
Anonymous
a b K Overclocking
October 5, 2004 11:20:54 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

David wrote:

> Michael Brown wrote:
>
> > Simon Hobbs wrote:
> >
> >>Howdie, I'm new.
> >>I put together a computer with a P4 2.8 It runs normally at about 40C,
> >>but as soon as it gets pushed, the temp rises quickly and the system
> >>freezes at about 47-50C.
> >
> >
> > How quickly is "quickly"? If it's reading the CPU die temperature, the
> > majority of the change should occur withing the first couple seconds after
> > the load is applied. If this does not occur, then you're reading the socket
> > temperature, which is more or less worthless in terms of determining actual
> > CPU temperature.

The temperature change is within the first through seconds. That said,
I think it must be hotter than the reading is telling me.

> >
> > [...]
> >
> >>My real qualm is, this is a Pentium, eh? Isn't it meant to pull back
> >>it's workrate to avoid over-heating.
> >
> >
> > The P4 should throttle to stop itself becoming *damaged* through
> > overheating.
>
> true.
>
> > However, intability usually occurs before damage occurs, so
> > it's quite possibly getting unstable (and thus crashing) before it reaches
> > the point where it starts to throttle.
>
> Well, it's supposed to throttle before it becomes unstable, it should just
> be 'slower', but that depends on the thermal solution being good enough to
> cool it when throttled.
>

I can only think there is something in that last statement. I should
improve the thermals first. I just can't see how it could heat so
quickly. I installed the chip, heat sync and fan as per instructions.
I've got one side of the case open and the room temperature is about
15 degrees. Maybe it isn't that hot and I have some bad RAM... Very
confusing.
Anonymous
a b K Overclocking
October 5, 2004 11:20:55 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

BananaOfTheNight wrote:

>> Well, it's supposed to throttle before it becomes unstable, it should
>> just be 'slower', but that depends on the thermal solution being good
>> enough to cool it when throttled.
>
>
> Actually, you don't need a cooling device to keep it stable when
> throttled down (at least for a few seconds) - have a look at this Tom's
> Hardware article and its associated video:
>
> http://www20.tomshardware.com/cpu/20010917/index.html

Yes, I've seen it, but I'm not sure that holds true through the entire
speed line and Intel certainly doesn't claim that it 'runs' with no
heatsink at all.

At any rate, with respect to the previous post saying the throttling was
only to prevent damage, here's what Intel says about it in the data sheets.

"Analysis indicates that real applications are unlikely to cause the
processor to consume maximum power dissipation for sustained periods of
time. Intel recommends that complete thermal solution designs target the
Thermal Design Power (TDP) indicated in Table 5-1 instead of the maximum
processor power consumption. The Thermal Monitor feature is intended to
help protect the processor in the unlikely event that an application
exceeds the TDP recommendation for a sustained period of time. For more
details on the usage of this feature, refer to Section 6.3. To ensure
maximum flexibility for future requirements, systems should be designed to
the Flexible Motherboard (FMB) guidelines, even if a processor with a lower
thermal dissipation is currently planned. In all cases, the Thermal Monitor
feature must be enabled for the processor to remain within specification."

So, even with a 'recommended' thermal solution, it is expected to throttle
under (rare) 'sustained full load' conditions and, of course, stay running.
Anonymous
a b K Overclocking
October 5, 2004 11:20:56 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Point taken.
Anonymous
a b K Overclocking
October 6, 2004 2:37:01 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

David Maynard wrote:
> BananaOfTheNight wrote:
>
>>> Well, it's supposed to throttle before it becomes unstable, it
>>> should just be 'slower', but that depends on the thermal solution
>>> being good enough to cool it when throttled.
>>
>>
>> Actually, you don't need a cooling device to keep it stable when
>> throttled down (at least for a few seconds) - have a look at this
>> Tom's Hardware article and its associated video:
>>
>> http://www20.tomshardware.com/cpu/20010917/index.html
>
> Yes, I've seen it, but I'm not sure that holds true through the entire
> speed line and Intel certainly doesn't claim that it 'runs' with no
> heatsink at all.

"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."
(Intel datasheet). I'm guessing a bare heatspreader counts as a
"significantly under-designed" thermal solution :) 

> At any rate, with respect to the previous post saying the throttling
> was only to prevent damage, here's what Intel says about it in the
> data sheets.
>
> "Analysis indicates that real applications are unlikely to cause the
> processor to consume maximum power dissipation for sustained periods
> of time. Intel recommends that complete thermal solution designs
> target the Thermal Design Power (TDP) indicated in Table 5-1 instead
> of the maximum processor power consumption. The Thermal Monitor
> feature is intended to help protect the processor in the unlikely
^^^^^^^^^^^^^^^^^^^^^^^^^^

In other words, it's to prevent damage to the CPU :)  However, I would say
that it's safe to assume that the CPU, under normal voltage and frequency
situations, should be stable, and not damaged, up to the rated maximum
temperature (or wherever the throttle is set to kick in). If the vcore or
FSB has been set to non-stock conditions though, or if the CPU is not being
supplied clean power, as would happen if you popped a Prescott into an older
board, then I would say that the chances of being unstable before hitting
the maximum temperature are certainly non-zero (though very hard to
quantify).

> event that an application exceeds the TDP recommendation for a
> sustained period of time. For more details on the usage of this
> feature, refer to Section 6.3. To ensure maximum flexibility for
> future requirements, systems should be designed to the Flexible
> Motherboard (FMB) guidelines, even if a processor with a lower
> thermal dissipation is currently planned. In all cases, the Thermal
> Monitor feature must be enabled for the processor to remain within
> specification."
>
> So, even with a 'recommended' thermal solution, it is expected to
> throttle under (rare) 'sustained full load' conditions and, of
> course, stay running.

--
Michael Brown
www.emboss.co.nz : OOS/RSI software and more :) 
Add michael@ to emboss.co.nz - My inbox is always open
Anonymous
a b K Overclocking
October 6, 2004 2:37:02 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Michael Brown wrote:

> David Maynard wrote:
>
>>BananaOfTheNight wrote:
>>
>>
>>>>Well, it's supposed to throttle before it becomes unstable, it
>>>>should just be 'slower', but that depends on the thermal solution
>>>>being good enough to cool it when throttled.
>>>
>>>
>>>Actually, you don't need a cooling device to keep it stable when
>>>throttled down (at least for a few seconds) - have a look at this
>>>Tom's Hardware article and its associated video:
>>>
>>>http://www20.tomshardware.com/cpu/20010917/index.html
>>
>>Yes, I've seen it, but I'm not sure that holds true through the entire
>>speed line and Intel certainly doesn't claim that it 'runs' with no
>>heatsink at all.
>
>
> "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."

Yeah. I knew I'd seen that in there but forgot where.

> (Intel datasheet). I'm guessing a bare heatspreader counts as a
> "significantly under-designed" thermal solution :) 

Hehe. I'd sure think so.


>>At any rate, with respect to the previous post saying the throttling
>>was only to prevent damage, here's what Intel says about it in the
>>data sheets.
>>
>>"Analysis indicates that real applications are unlikely to cause the
>>processor to consume maximum power dissipation for sustained periods
>>of time. Intel recommends that complete thermal solution designs
>>target the Thermal Design Power (TDP) indicated in Table 5-1 instead
>>of the maximum processor power consumption. The Thermal Monitor
>>feature is intended to help protect the processor in the unlikely
>
> ^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> In other words, it's to prevent damage to the CPU :) 

I figured you'd pick up on that ;)  But it's a matter of 'reading' like
engineers 'write'. It 'protects' the processor by keeping it in the safe
operating zone.

Otherwise you'd have to interpret it as saying "well, we expect it to crash
if you do something unusually processor intensive." <g>

> However, I would say
> that it's safe to assume that the CPU, under normal voltage and frequency
> situations, should be stable, and not damaged, up to the rated maximum
> temperature (or wherever the throttle is set to kick in).

Right. Both, actually. But with a proper thermal solution, in combination
with the throttle, it shouldn't ever reach the 'rated maximum'.

> If the vcore or
> FSB has been set to non-stock conditions though, or if the CPU is not being
> supplied clean power, as would happen if you popped a Prescott into an older
> board, then I would say that the chances of being unstable before hitting
> the maximum temperature are certainly non-zero (though very hard to
> quantify).

Oh, definitely. And the same goes for overclocking. Might; might not. You
can't use a spec sheet to tell you what it 'will do' when you operate it
outside the specs.

Beginning overclockers make that mistake a lot. Casually noting "data sheet
says it's good to 70C" but they've got Vcore cranked up past even the
static, non operating, maximum.


>>event that an application exceeds the TDP recommendation for a
>>sustained period of time. For more details on the usage of this
>>feature, refer to Section 6.3. To ensure maximum flexibility for
>>future requirements, systems should be designed to the Flexible
>>Motherboard (FMB) guidelines, even if a processor with a lower
>>thermal dissipation is currently planned. In all cases, the Thermal
>>Monitor feature must be enabled for the processor to remain within
>>specification."
>>
>>So, even with a 'recommended' thermal solution, it is expected to
>>throttle under (rare) 'sustained full load' conditions and, of
>>course, stay running.
>
>
> --
> Michael Brown
> www.emboss.co.nz : OOS/RSI software and more :) 
> Add michael@ to emboss.co.nz - My inbox is always open
>
>
Anonymous
a b K Overclocking
October 8, 2004 10:03:49 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

get_simon_hobbs@hotmail.com (Simon Hobbs) wrote in message news:<ec1a72b1.0410040334.77c0dc2e@posting.google.com>...
> Howdie, I'm new.
> I put together a computer with a P4 2.8 It runs normally at about 40C,
> but as soon as it gets pushed, the temp rises quickly and the system
> freezes at about 47-50C. I don't run games or anything hardcore, I
> first realised why it was happening when I watched the anti-virus
> program do a full system scan. But this is hardly gruntwork, the CPU
> runs at about 50% for 5 minutes.
>
> Now I'm willing to accept the temp monitors (PC Alert 4) might not be
> accurate, maybe it's running hotter because of my amateur
> box-building. So from reading other posts I reckon I should buy some
> gluey stuff for the heat sync and maybe put a proper fan in and all
> that newbie stuff - OK I can give that a go.
>
> My real qualm is, this is a Pentium, eh? Isn't it meant to pull back
> it's workrate to avoid over-heating. While I'm here asking
> overclockers (who better to ask), I'm not actually overclocking
> myself. So is there something I've missed? Have I got a dud?
>
> And is it damaging the cpu? Or is the cpu shutting itself down before
> the damage?
>
> Would love your say, I say, .s
> (MSI 865GM2 mobo, nvidia geforce2)

Yep, I *was* green and it *did* freeze quickly. I blame my ineptitude
on two things: having the IKEA allen-key mentality and putting
together PCs once every olympiad. No, what the hell, I say Intel
should do a hack's DYI version of instruction where they put a little
more emphasis on the importance of that pasty conduction stuff. I had
a nagging feeling that silver-blackish material was important for
something...

All right. For those other newbies who follow in my terrible wake:

***Go to a computer or electronics shop and buy some "heat conduction
paste". The computer shop will sell more expensive stuff that works
better. Every time you take the fan and header spreader off, re-apply
the paste.***

Ah, I feel tremendous. I've been promoted from seriously-under-clocked
to normal-guy-with-computer. And in the process, i learnt how to check
my ram, monitor the voltage to the CPU and calculate prime numbers.
So, now that my wife can check her emails without the computer
crashing, I think I'll get back to the garden...

Big thanks to those who responded to my OP with guidance.
Simon
!