CPU temps stable, then rising w/o load?

[S. Whitmore]

I recently upgraded my system and now have a new AMD Athlon 64 3200+ on
a new ASUS K8V SE Deluxe mobo, with a new Enermax 470W "Noisetaker" PSU.

Ambient temps have been consistenly in the 65-75F range, and for the
couple of weeks I've had the new hardware running, the mobo and CPU
temps have been pretty consistently in the 75-85F range (depending on
load, of course, and lately I haven't been doing much as I also
reorganized my hard drives and I'm still getting things reinstalled and
reconfigured). Those temps are as reported by ASUS PC Probe, which I
have running all the time in the background (i.e., tooltray).

Tonight I got the unpleasant surprise of my system shutting down
suddenly. No warning, no error, no alarm or alert from PC Probe, no
gentle close -- just suddenly a dark screen, silence, and the monitor
shifting into sleep mode. I powered it up again and it ran for a short
time and then it shut down again, just as abruptly. Both times, the
Ethernet indicators were still lit on the mobo and the PSU was doing
it's post-shutdown cool-down (this PSU continues exhausting the case for
a couple minutes after shutdown).

My first "suspect" was a power problem, specifically with the UPS,
because another of the same type that I got in the same order from
newegg.com just went back w/ an RMA today. But it didn't really seem
right, so I did some looking online in various discussion groups. Most
pointed to either a PSU failure (possible) or a temp problem. I didn't
think it would be temp related because the system wasn't under any real
load and hadn't been on very long. The first time it happened, I was in
the middle of rebooting after changing my network configuration. The
second time it happened I was just reading a message in Mozilla mail.

Despite thinking it wasn't a temp problem, I decided it was easy to
check for that, so I booted the system and went right into the BIOS
Setup. Watched, and didn't like what I saw. The mobo temp stayed in
the same range it's been in, but the CPU was steadily climbing as I
watched. When it got to 110F (after about 3 or 4 minutes) I shut it
down, since I was already seeing what looked to me like bad behavior.
(I'm assuming sitting and watching the hardware monitor in the BIOS
Setup does not impose much of a CPU load -- am I wrong?)

After awhile, I decided to ask online, so I booted up to Windows 2000.
As I write this message, PC Probe is again saying that my CPU temp is in
the 75-85F range (at the moment, 28C/82F). So maybe the problem has
gone away for now, or maybe PC Probe isn't reporting temps correctly.

I would appreciate ideas and input on this abrupt shutdown problem
and/or the CPU temp behavior that I'm seeing.

TIA,

 

[John Doe]

....

The first time it happened, I was in the middle of rebooting after
changing my network configuration. The second time it happened I
was just reading a message in Mozilla mail.

I think that's your suspect (network, and input or video stuff). As
a nuke style workaround, you could try switching from Windows NT to
Windows XP, if possible.

After awhile, I decided to ask online, so I booted up to Windows
2000. As I write this message, PC Probe is again saying that my CPU
temp is in the 75-85F range (at the moment, 28C/82F).

It's like Hawaii inside your case.

So maybe the problem has gone away for now, or maybe PC Probe isn't
reporting temps correctly.

Because it's not over 85°F? Where did you get your information about
CPU temperature, so I can avoid it.

I would appreciate ideas and input on this abrupt shutdown problem
and/or the CPU temp behavior that I'm seeing.

My CPU temperature varies significantly with room temperature.

You are overconcerned about CPU temperature. Next time, when it gets
to 43°C, do not shut off your computer. Too hot for my CPU, in my
opinion, is over about 52°C. The maximum operating temperature of an
Athlon XP CPU is about 85°C which is 185°F.

Good luck.

 

[S. Whitmore]

I think that's your suspect (network, and input or video stuff).

To cause an abrupt shutdown? I'd consider that doubtful at best, based
on my experience.

As
a nuke style workaround, you could try switching from Windows NT to
Windows XP, if possible.

Nope, not going to XP. Linux maybe, but not XP. Windows 2000 has been
the most stable Microsoft OS I've used (well, other than MS-DOS 3.3
maybe), so I have no incentive at the moment to switch to something that
adds the annoyance of "activation" just to play around with my system.

It's like Hawaii inside your case.

Ah, so the hula girls caused the shutdown. Figures!

You are overconcerned about CPU temperature.

No, I'm concerned about having my system shut off abruptly, without
warning from either software tool I expected to warn me (i.e., the UPS
monitoring software for power failures and ASUS PC Probe for temp or fan
issues).

Next time, when it gets
to 43°C, do not shut off your computer. Too hot for my CPU, in my
opinion, is over about 52°C. The maximum operating temperature of an
Athlon XP CPU is about 85°C which is 185°F.

Well, I didn't shut it down for being too hot. I shut it down because I
was already seeing behavior that I didn't expect based on what I've
observed in the week or two that I've had this upgraded system
operational, and didn't want to sit and watch until it did climb too
high (which, of course, was based on the assumption that it would
continue rising). I wouldn't get concerned about the temperature itself
until it was over 55C, based on what I've read about this chip (where
60C seems to be a key level). Sorry if I gave the wrong impression.

 

[John Doe]

....

To cause an abrupt shutdown?

Yes. As I recall, shutting down happens.
These days, saying "shut down Windows" can do it.

I'd consider that doubtful at best, based on my experience.

I'm having trouble reading between the lines.

Your entire original post dwelled on the idea that it was CPU
overheating. Given the facts, I can't imagine a worse lead.

Nope, not going to XP. Linux maybe, but not XP. Windows 2000 has
been the most stable Microsoft OS I've used (well, other than
MS-DOS 3.3 maybe), so I have no incentive at the moment to switch
to something that adds the annoyance of "activation" just to play
around with my system.

There is an easy workaround.

No, I'm concerned about having my system shut off abruptly, without
warning from either software tool I expected to warn me (i.e., the
UPS monitoring software for power failures and ASUS PC Probe for
temp or fan issues).

I think it's a low-level software problem, but maybe you could test
for that one if your BIOS allows an after power failure state of ON.




....

 

[Mxsmanic]

I recently upgraded my system and now have a new AMD Athlon 64 3200+ on
a new ASUS K8V SE Deluxe mobo, with a new Enermax 470W "Noisetaker" PSU.

But you left out the most important part: What OPERATING SYSTEM are you
running?

Some older versions of Windows (those not derived from Windows NT) don't
actually halt the processor when they are idle; instead, they run a
continuous "idle loop" that keeps the processor pegged at 100%. Some
other older operating systems work the same way.

More modern and better-written operating systems actually halt the
processor when there is nothing to do. This is true for anything based
on UNIX (including Mac OS X) and anything based on Windows NT (Windows
NT itself, Windows 200x, and Windows XP). I don't know if the old Mac
OS did this.

If your machine is running an operating system with an "idle loop,"
temperatures will climb to just about the maximum even when the system
seems quiet.

 

[Mxsmanic]

Nope, not going to XP.

XP, 2000, and NT all have the same kernel and code base, so switching
from one to another would make no difference in this case.

 

[John Doe]

S. Whitmore writes:

XP, 2000, and NT all have the same kernel and code base,

That is false.

so switching to from one to another would make no difference in
this case.

That is wild speculation based on faulty information or at best
based on a half truth.

I'm not trying to sell Microsoft products. I'm simply trying to help
the user solve his problem. Please keep in mind I called the
suggestion about switching to Windows XP a "nuke style workaround".

In fact, Windows XP is a more modern operating system which handles
things better, whether or not you can cope with that idea. It was an
uneasy upgrade for me as well.

 

[David Maynard]

I recently upgraded my system and now have a new AMD Athlon 64 3200+ on
a new ASUS K8V SE Deluxe mobo, with a new Enermax 470W "Noisetaker" PSU.

Ambient temps have been consistenly in the 65-75F range, and for the
couple of weeks I've had the new hardware running, the mobo and CPU
temps have been pretty consistently in the 75-85F range (depending on
load, of course, and lately I haven't been doing much as I also
reorganized my hard drives and I'm still getting things reinstalled and
reconfigured). Those temps are as reported by ASUS PC Probe, which I
have running all the time in the background (i.e., tooltray).

Tonight I got the unpleasant surprise of my system shutting down
suddenly. No warning, no error, no alarm or alert from PC Probe, no
gentle close -- just suddenly a dark screen, silence, and the monitor
shifting into sleep mode. I powered it up again and it ran for a short
time and then it shut down again, just as abruptly. Both times, the
Ethernet indicators were still lit on the mobo and the PSU was doing
it's post-shutdown cool-down (this PSU continues exhausting the case for
a couple minutes after shutdown).

My first "suspect" was a power problem, specifically with the UPS,
because another of the same type that I got in the same order from
newegg.com just went back w/ an RMA today. But it didn't really seem
right, so I did some looking online in various discussion groups. Most
pointed to either a PSU failure (possible) or a temp problem. I didn't
think it would be temp related because the system wasn't under any real
load and hadn't been on very long. The first time it happened, I was in
the middle of rebooting after changing my network configuration. The
second time it happened I was just reading a message in Mozilla mail.

Despite thinking it wasn't a temp problem, I decided it was easy to
check for that, so I booted the system and went right into the BIOS
Setup. Watched, and didn't like what I saw. The mobo temp stayed in
the same range it's been in, but the CPU was steadily climbing as I
watched. When it got to 110F (after about 3 or 4 minutes) I shut it
down, since I was already seeing what looked to me like bad behavior.
(I'm assuming sitting and watching the hardware monitor in the BIOS
Setup does not impose much of a CPU load -- am I wrong?)

Yes, because the CPU does not go into any power saving or sleep modes when
in the BIOS like it does when the operating system is at idle; which is
most of the time if you're not doing anything intensive.

After awhile, I decided to ask online, so I booted up to Windows 2000.
As I write this message, PC Probe is again saying that my CPU temp is in
the 75-85F range (at the moment, 28C/82F). So maybe the problem has
gone away for now, or maybe PC Probe isn't reporting temps correctly.

It's because the O.S. puts it in sleep mode when idle.

I would appreciate ideas and input on this abrupt shutdown problem
and/or the CPU temp behavior that I'm seeing.

The shutdown could be caused by a myriad of things. Check your event logs,
although your symptoms suggest it dies before an event could be recorded.

 

[Mxsmanic]

That is false.

I've read the source code.

That is wild speculation based on faulty information or at best
based on a half truth.

See above.

Current Microsoft Windows operating systems are derived from the Windows
NT code base, because NT is very well written. The base has evolved and
differs somewhat from product to product, and there are cosmetic and
configuration differences, but under the hood they are mostly the same,
as careful examination of the operating systems reveals.

 

[Matt]

S. Whitmore writes:




But you left out the most important part: What OPERATING SYSTEM are you
running?

Some older versions of Windows (those not derived from Windows NT) don't
actually halt the processor when they are idle; instead, they run a
continuous "idle loop" that keeps the processor pegged at 100%. Some
other older operating systems work the same way.

More modern and better-written operating systems actually halt the
processor when there is nothing to do. This is true for anything based
on UNIX (including Mac OS X) and anything based on Windows NT (Windows
NT itself, Windows 200x, and Windows XP). I don't know if the old Mac
OS did this.

If your machine is running an operating system with an "idle loop,"
temperatures will climb to just about the maximum even when the system
seems quiet.

What is the typical behavior of BIOSes at idle?

I had a Biostar M7NCD Pro with Athlon XP 2500+ (Barton). I normally run
Linux, but I don't have a way to check CPU temp under Linux. But when
it was idling in Linux and I shut Linux down and went immediately
(without power down) into the health-status page of the BIOS, I would
see the temp rise in about twenty minutes from about 38 C to about 48 C.

So it seemed to run about 10 C hotter when idling in the BIOS than when
idling in Linux.

 

[JAD]

a troll disagrees with you in one line nonsense.


Path:
be02_text!hwmnpeer02.lga!hw-filter.lga!hwmnpeer01.lga!hwmedia!cycny01.gnilin
k.net!cyclone1.gnilink.net!gnilink.net!wns14feed!worldnet.att.net!207.115.63
..142!newscon02.news.prodigy.com!prodigy.net!newsmst01a.news.prodigy.com!prod
igy.com!postmaster.news.prodigy.com!newssvr11.news.prodigy.com.POSTED!674205
56!not-for-mail
Newsgroups: alt.comp.hardware.pc-homebuilt
Subject: Re: CPU temps stable, then rising w/o load?
From: John Doe <[email protected]>
References: <FcEOd.16937$uc.8354@trnddc04>
<[email protected]> <j6FOd.16206$uc.1566@trnddc08>
<[email protected]>
Organization: sometimes
Message-ID: <[email protected]>
User-Agent: Xnews/06.08.25
Lines: 59
NNTP-Posting-Host: 69.153.49.208
X-Complaints-To: (e-mail address removed)
X-Trace: newssvr11.news.prodigy.com 1108030519 ST000 69.153.49.208 (Thu, 10
Feb 2005 05:15:19 EST)
NNTP-Posting-Date: Thu, 10 Feb 2005 05:15:19 EST
X-UserInfo1:
TSU[@SVDZZWARWD[FJKD^RLARJT@QDDMEPWXODMMHXMTWA]EPMS[AACY@TZZXQ[KS^ESKCJLOF_J
_NGAWNTG^_XGTNTAHULK[X[NRTC@G\P^PLT_OCBRHUO@@TBQZDZMHD[YZ@NLXQXIWMOSXT_KOLK^
^CXFF\WHMI^C@EGA_[FXAQ@E^TGNMUXGYNS[QQVL
Date: Thu, 10 Feb 2005 10:15:19 GMT
Xref: Hurricane-Charley alt.comp.hardware.pc-homebuilt:39325

S. Whitmore writes:

XP, 2000, and NT all have the same kernel and code base,

That is false.

so switching to from one to another would make no difference in
this case.

That is wild speculation based on faulty information or at best
based on a half truth.

I'm not trying to sell Microsoft products. I'm simply trying to help
the user solve his problem. Please keep in mind I called the
suggestion about switching to Windows XP a "nuke style workaround".

In fact, Windows XP is a more modern operating system which handles
things better, whether or not you can cope with that idea. It was an
uneasy upgrade for me as well.

 

[Mxsmanic]

What is the typical behavior of BIOSes at idle?

A continuous loop, from what I've seen. Even if you do nothing in the
BIOS menus, the temperature of the processor will gradually rise as it
loops waiting for input.

I had a Biostar M7NCD Pro with Athlon XP 2500+ (Barton). I normally run
Linux, but I don't have a way to check CPU temp under Linux. But when
it was idling in Linux and I shut Linux down and went immediately
(without power down) into the health-status page of the BIOS, I would
see the temp rise in about twenty minutes from about 38 C to about 48 C.

So it seemed to run about 10 C hotter when idling in the BIOS than when
idling in Linux.

Linux halts the processor when the OS is idle. The BIOS doesn't.

 

[JAD]

it seems to me that you have all the hardware angles covered, less the Hard
drive failure scenario. You never mentioned Antivirus. 2000 has its exploits
too. As your probably aware, there were many virus patterns that included a
warning then shut down and then some that just shut you down.

Your temps are fine IMO.

 

[Matt]

a troll disagrees with you in one line nonsense.


Path:
be02_text!hwmnpeer02.lga!hw-filter.lga!hwmnpeer01.lga!hwmedia!cycny01.gnilin
k.net!cyclone1.gnilink.net!gnilink.net!wns14feed!worldnet.att.net!207.115.63
.142!newscon02.news.prodigy.com!prodigy.net!newsmst01a.news.prodigy.com!prod
igy.com!postmaster.news.prodigy.com!newssvr11.news.prodigy.com.POSTED!674205
56!not-for-mail
Newsgroups: alt.comp.hardware.pc-homebuilt
Subject: Re: CPU temps stable, then rising w/o load?
From: John Doe <[email protected]>
References: <FcEOd.16937$uc.8354@trnddc04>
<[email protected]> <j6FOd.16206$uc.1566@trnddc08>
<[email protected]>
Organization: sometimes
Message-ID: <[email protected]>
User-Agent: Xnews/06.08.25
Lines: 59
NNTP-Posting-Host: 69.153.49.208
X-Complaints-To: (e-mail address removed)
X-Trace: newssvr11.news.prodigy.com 1108030519 ST000 69.153.49.208 (Thu, 10
Feb 2005 05:15:19 EST)
NNTP-Posting-Date: Thu, 10 Feb 2005 05:15:19 EST
X-UserInfo1:
TSU[@SVDZZWARWD[FJKD^RLARJT@QDDMEPWXODMMHXMTWA]EPMS[AACY@TZZXQ[KS^ESKCJLOF_J
_NGAWNTG^_XGTNTAHULK[X[NRTC@G\P^PLT_OCBRHUO@@TBQZDZMHD[YZ@NLXQXIWMOSXT_KOLK^
^CXFF\WHMI^C@EGA_[FXAQ@E^TGNMUXGYNS[QQVL
Date: Thu, 10 Feb 2005 10:15:19 GMT
Xref: Hurricane-Charley alt.comp.hardware.pc-homebuilt:39325

S. Whitmore writes:


XP, 2000, and NT all have the same kernel and code base,

That is false.

so switching to from one to another would make no difference in
this case.

That is wild speculation based on faulty information or at best
based on a half truth.

I'm not trying to sell Microsoft products. I'm simply trying to help
the user solve his problem. Please keep in mind I called the
suggestion about switching to Windows XP a "nuke style workaround".

In fact, Windows XP is a more modern operating system which handles
things better, whether or not you can cope with that idea. It was an
uneasy upgrade for me as well.

That is false.



That is wild speculation based on faulty information or at best
based on a half truth.

I'm not trying to sell Microsoft products. I'm simply trying to help
the user solve his problem. Please keep in mind I called the
suggestion about switching to Windows XP a "nuke style workaround".

In fact, Windows XP is a more modern operating system which handles
things better, whether or not you can cope with that idea. It was an
uneasy upgrade for me as well.






newssvr11.news.prodigy.com!newscon03.news.prodigy.com!newsmst01a.news.prodig
y.com!prodigy.com!newscon06.news.prodigy.com!prodigy.net!border1.nntp.dca.gi
ganews.com!nntp.giganews.com!local1.nntp.dca.giganews.com!news.giganews.com.
POSTED!not-for-mail


sv3-qVijxNYX4GdNN0+krugUcV3UB0ZtsG8jU/f8ZhcCNuBKeBzRlwbNPqVFg4G9/Z+byz3dD8eo
bx08Mze!wVbM08iTYGwdH4U+E9REQCcXucQkX3Px/f0+VjlUATATHfD6JFVDEWwAgiBftsojlA==

complaint properly

LOL

 

[Matt]

Matt writes:




A continuous loop, from what I've seen. Even if you do nothing in the
BIOS menus, the temperature of the processor will gradually rise as it
loops waiting for input.




Linux halts the processor when the OS is idle. The BIOS doesn't.

Thanks. Presumably the main or only advantage of halting is to save energy.

 

[John Doe]

A not quite grown-up perpetually top posting troll who cannot handle
being corrected for giving obviously bad advice to someone seeking
help from this group.

http://tinyurl.com/3lneg

Message-ID: <[email protected]>
Message-ID: <8MABd.47038$pF1.42093 @fe06.lga>

Path: newssvr12.news.prodigy.com!newsdbm05.news.prodigy.com!newsdst02.news.prodigy.com!newsmst01a.news.prodigy.com!prodigy.com!newscon06.news.prodigy.com!prodigy.net!news-feed01.roc.ny.frontiernet.net!nntp.frontiernet.net!news.glorb.com!atl-c03.usenetserver.com!hwmnpeer01.phx!hwmedia!hw-poster!fe06.lga.POSTED!53ab2750!not-for-mail
From: "JAD" <[email protected]>
From: "JAD" <kapasitor @earthcharter.net>
Newsgroups: alt.comp.hardware.pc-homebuilt
References: <FcEOd.16937$uc.8354@trnddc04> <[email protected]> <j6FOd.16206$uc.1566@trnddc08> <[email protected]> <[email protected]>
Subject: Re: CPU temps stable, then rising w/o load?
Lines: 126
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2800.1409
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
Message-ID: <[email protected]>
Message-ID: <WDPOd.4997$dZ.3890 @fe06.lga>
X-Trace: dkoicekneficefklldeombdpnajfmnhoamgkolminjdchkjkfcccdnnagloeleneplnamjofdhaonobpekbcabnehelknhebnnfbcinncalfhmhfkpcfpjhbbdnejfmhegcmjhhibopfakdb
NNTP-Posting-Date: Thu, 10 Feb 2005 13:39:18 MST
Date: Thu, 10 Feb 2005 12:33:39 -0800
Xref: newsmst01a.news.prodigy.com alt.comp.hardware.pc-homebuilt:429117

a troll disagrees with you in one line nonsense.

....

 

[S. Whitmore]

Yes, because the CPU does not go into any power saving or sleep modes
when in the BIOS like it does when the operating system is at idle;
which is most of the time if you're not doing anything intensive.

Ok, thanks -- this is the most salient point of this thread. So I will
continue looking at reasons other than a steadily rising "at-rest" CPU
temperature since that wasn't what I was seeing after all (i.e., it was
not "at rest").

The shutdown could be caused by a myriad of things. Check your event
logs, although your symptoms suggest it dies before an event could be
recorded.

Yeah, I haven't looked at the system logs yet but I doubt there's
anything there, it's too abrupt of a shutdown; also, the first shutdown
was before any OS was loaded, which would preclude any OS-related logs.
Given other things I've read online, the PSU is back to being my #1
suspect, although there are a couple other things I could see playing a
role.

Thanks,

 

[S. Whitmore]

Yes. As I recall, shutting down happens.
These days, saying "shut down Windows" can do it.

Not when Windows isn't running, i.e., when the system abruptly shuts
down before the OS is loaded.

I'm having trouble reading between the lines.

Your entire original post dwelled on the idea that it was CPU
overheating. Given the facts, I can't imagine a worse lead.

Sorry the logic of it is escaping you. First, understand that an
overheated CPU (at least, of the type relevant here, per my original
post) will shut down. I've never experienced that, so I don't know what
that will "look like" -- i.e., the system as a whole shutting down,
or...? Second, understand that I was accustomed to seeing my CPU temps
in a steady range when I looked at it after Windows 2000 was up and
running, using ASUS PC Probe, but when I was looking at it in the BIOS,
the temperature seemed to be rising steadily without my doing anything
-- and that I didn't realize at that time that the BIOS would keep the
CPU busy even when I wasn't doing anything.

So given the facts (as I understood them at the time) -- that an
overheated CPU will shut down and that the CPU is "mysteriously" heating
up when it "shouldn't" be -- I can't imagine why you can't imagine a
worse lead.

Now that I know there's no mystery in the rising CPU temperature, I'll
be looking at other factors. Probably the PSU, since it's highly
doubtful that it's related to Windows 2000, since the first shut-down
happened before Windows 2000 even started to load. (In my original post
I indicated that "[t]he first time it happened, I was in the middle of
rebooting after changing my network configuration" but I could have been
more clear by pointing out that the reboot process had not proceeded as
far as Windows 2000 starting to load.)

 

[S. Whitmore]

But you left out the most important part: What OPERATING SYSTEM are you
running?

Sorry, I was not clear enough about a certain point in my original post.
When I said that "[t]he first time it happened, I was in the middle of
rebooting after changing my network configuration," I should have made
it clear that it was before any OS started to load. Thus, I was not
thinking about OS-specific causes, and only looking at the hardware.

FWIW, it's a dual boot system with Windows 2000 and GNU/Linux (Slackware
10). When it crashed, I'm fairly sure that I had not even seen the LILO
prompt. (Although, to be fair, I wasn't expecting this to happen so I
wasn't watching very closely -- however, I do recall seeing the ASUS
boot logo on the screen right before or as the system shut down.)

Thanks,

 

[John Doe]

John Doe writes:

I've read the source code.

Hey ya, Burt!

See above.
Current Microsoft Windows operating systems are derived from the
Windows NT code base, because NT is very well written. The base
has evolved and differs somewhat from product to product, and there
are cosmetic and configuration differences, but under the hood they
are mostly the same, as careful examination of the operating
systems reveals.

I guess you could write volumes on the subject. Who knows.

Your dismissal of my suggestion to upgrade the operating system is
not based on the problem as described.

The original poster recently upgraded to a new mainboard and CPU.
Like it or not, new hardware potentially benefits from a more recent
operating system, probably/typically more than any other change in
circumstance. Windows XP produces fewer failures. Windows XP is the
current technology. Windows 2000 and Windows NT are aging.