On Oct 30, 2:44 am, Paul <nos...@needed.com> wrote:
>
> Test with Prime95 as a load test, and check the temperature when
> that is running.
>
> Check that the cooler is making good contact, and there is sufficient
> thermal paste.
>
> A Prescott P4 will run hotter than a Northwood P4. The Prescott could
> have a 1MB cache, while the Northwood would have a 512KB cache. (There
> are things like Extreme processors, that may have other cache configurations.)
>
> Use the Intel utility to identify your processor.
>
> http://www.intel.com/support/process.../CS-014921.htm
>
> Paul
This is the P4E - 800
1 P43000E478 -- Intel Pentium 4 3EGHz 800MHz 1MB Socket 478 CP...
$7.99 $7.99
Good idea - Prime95. Seems there's 3 levels of cache 16K-1M,
something in between. Thought I had a killer CoolerMaster 3-pipe
heatsink setup, it's sure a monster, but when I took it out of the
package and started looking to fit adaptors together, I must've
misread a sale for ordering for a 478 setup. AMD2/3 and 775. Will
have to use it on something else or consider doing a "modification."
So in between trying to figure why I'd missed the boat on that order,
I put back on my old Intel 478 heatsink that came with the Celeron D.
Two layers of compound. Maybe a little thin on the heatsink, but a
nice spread over the P4. Seems it went in a little odd, first the P4
had a rocking motion I couldn't get rid of before locking down the
socket lever, heard a weird snap when tightening down the fan levers
onto the CPU shroudings, and light wasn't right either while working
inside an assembled case. Wouldn't hurt to take it all off and get
into it more carefully for a redo. I'd expected some P4 heat for
85watts, ran into others saying besides hot, they'll fry without much
provocation by bumping the host freq. It's totally unlike a 45watt
AMD Orleans I also did last week. Thought something wrong with its
sensors - 88F all the time. Went to 93F, though, after pulled some
fans out of the case. Clean without much dirt in the air going
through, and strangely silent.
There's nothing about that P4-E reading that isn't other than some
sort of strange glitch. I mean when I first saw 162F I was in the
case in a couple of seconds, grabbing the CPU cooler, expecting it to
fall off or something. Think the BIOS reported it had bumped up the
CPU fan from 2000 to 3000rpm over four, five hours, but two audio amps
beside it were the same if not hotter than that CPU heatsink. (Have
an IR temperature gun). It's got to be between a hard-to-software
transition since the BIOS immediately was having nothing to do with
that nonsense. Curious what ASUS Probe would say if it happens
again. Could try some mandel fractals, a string of audio files being
normalized, a demanding video encode boxed, and Prime 95 iterations.
I'm used to 125F tops with AMD when they're on their knees and going
borderline unstable. Intel I'd expect can be pushed harder.
Mismatched 512M mem modules (Mushkin/Crucial) seems my only wildcard
outside of trying to stay stock by "auto/default" BIOS settings.
Linear Mode
