CPU heat?

[Deleted member k]

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

I upgraded my CPU last night from a P4 1.6 GHz
to a P4 2.4 GHz. The motherboard is a Soyo P41 Fire Dragon. The Bios, AIDA32
and Smartguardian reported an average temp. of 45C. on the old P4 1.6. The
new CPU is reported at 48C. from the Bios, but 78C. from AIDA32 and
Smartguardian! I used Arctic Silver II. Any chance I'm running that hot? I
also ran 3D Mark 2001 before switching CPU's. I rebooted to the desktop
before each test and ran the benchmark with my normal Win XP Pro setup.
That's all services I normally run plus ZA, AntiVirus, and wireless network.
I was looking for a comparison of both processors and not a high score. The
results averaged 7090 for the P4 1.6 and only 7920 for the P4 2.4. Not much
of an increase. Any thoughts on the temp. and 3D Mark?

512 MB of Ram
AIW 9600

K

 

[Guest]

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

"K" <name@swbell.net> writes:
> I upgraded my CPU last night from a P4 1.6 GHz
>to a P4 2.4 GHz. The motherboard is a Soyo P41 Fire Dragon. The Bios, AIDA32
>and Smartguardian reported an average temp. of 45C. on the old P4 1.6. The
>new CPU is reported at 48C. from the Bios, but 78C. from AIDA32 and
>Smartguardian! I used Arctic Silver II. Any chance I'm running that hot?

Yow! I'd try to find some thermometer not running on software to check.
Radio Shack had a cute fairly cheap infra-red thermometer, you aimed it
at something and read out the temperature. A shiny aluminum heatsink
might read less than a flat black anodized surface but I'd think getting
"a second, well third, opinion" would be a good idea before running long.

>I also ran 3D Mark 2001 before switching CPU's. I rebooted to the desktop
>before each test and ran the benchmark with my normal Win XP Pro setup.
>That's all services I normally run plus ZA, AntiVirus, and wireless network.
>I was looking for a comparison of both processors and not a high score. The
>results averaged 7090 for the P4 1.6 and only 7920 for the P4 2.4. Not much
>of an increase. Any thoughts on the temp. and 3D Mark?

I congratulate you for doing before and after measurements, most don't
and I'm impressed that you did. It would be really nice if people and
magazine articles would tell you you were going to get x% increase for
doing what they describe. And less than 20% often is invisible unless
you have something to measure it with.

It isn't surprising to compare something with even 60% higher clock rate
and find less than 20% increase in speed. Those were the numbers I found
when I did my timings. You got 12% for a 50% increase, not wildly out
of line. I suspect that in the chain of hardware increasing the speed
makes one of the items mostly the bottleneck.

>512 MB of Ram
>AIW 9600

I hope something I wrote here was useful to you.

 

[alex]

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

What graphics card you using? Chances are the very little increase is due
to the video card being the bottleneck, not the processor. What does MBM say
about the temps?

 

[Deleted member k]

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

As noted in the OP it's a ATI AIW 9600. Have not tried MBM but will do so
tonight>
K

"Alex" <Freeflyer91@hotmail.com.NOSPAM> wrote in message
news:c9nuvl$f9f$1@news7.svr.pol.co.uk...
> What graphics card you using? Chances are the very little increase is due
> to the video card being the bottleneck, not the processor. What does MBM
say
> about the temps?
>
>

 

[Deleted member k]

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

"Don Taylor" <dont@agora.rdrop.com> wrote in message
news:erOdnWub27--MyLdRVn-sA@scnresearch.com...
> "K" <name@swbell.net> writes:
> > I upgraded my CPU last night from a P4 1.6 GHz
> >to a P4 2.4 GHz. The motherboard is a Soyo P41 Fire Dragon. The Bios,
AIDA32
> >and Smartguardian reported an average temp. of 45C. on the old P4 1.6.
The
> >new CPU is reported at 48C. from the Bios, but 78C. from AIDA32 and
> >Smartguardian! I used Arctic Silver II. Any chance I'm running that hot?
>
> Yow! I'd try to find some thermometer not running on software to check.
> Radio Shack had a cute fairly cheap infra-red thermometer, you aimed it
> at something and read out the temperature. A shiny aluminum heatsink
> might read less than a flat black anodized surface but I'd think getting
> "a second, well third, opinion" would be a good idea before running long.
>
> >I also ran 3D Mark 2001 before switching CPU's. I rebooted to the desktop
> >before each test and ran the benchmark with my normal Win XP Pro setup.
> >That's all services I normally run plus ZA, AntiVirus, and wireless
network.
> >I was looking for a comparison of both processors and not a high score.
The
> >results averaged 7090 for the P4 1.6 and only 7920 for the P4 2.4. Not
much
> >of an increase. Any thoughts on the temp. and 3D Mark?
>
> I congratulate you for doing before and after measurements, most don't
> and I'm impressed that you did. It would be really nice if people and
> magazine articles would tell you you were going to get x% increase for
> doing what they describe. And less than 20% often is invisible unless
> you have something to measure it with.
>
> It isn't surprising to compare something with even 60% higher clock rate
> and find less than 20% increase in speed. Those were the numbers I found
> when I did my timings. You got 12% for a 50% increase, not wildly out
> of line. I suspect that in the chain of hardware increasing the speed
> makes one of the items mostly the bottleneck.
>
> >512 MB of Ram
> >AIW 9600
>
> I hope something I wrote here was useful to you.

It was and I appreciate the comments.
K

 

[Guest]

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

Upgrade a CPU when your application in CPU bound (*), most are
not and often a hard drive upgrade from older to modern is better.

(*) where a FSB goes from 400 to 800 is another factor of course.

o Check the CPU heatsink is seated correctly
o Use a TINY amount of thermal compound
---- it is not butter, spread a tiny droplet out well
---- two metal surfaces have tiny voids - full or air
---- air is a poor conductor even in such tiny voids
---- thermal compound is just to fill those tiny voids

Thermal compound merely transfer heat better than the air.
Adding more compound doesn't get heat out better or such
--
Dorothy Bradbury
www.stores.ebay.co.uk/panaflofan for quiet fans, books & other items
http://homepage.ntlworld.com/dorothy.bradbury/panaflo.htm (Direct)