CPU temp - Suse

This is a discussion on CPU temp - Suse ; Hans-Peter Diettrich wrote: > Right, but steam from a wet finger indicates > 100C ;-) So does a chip logo branded on your pinky. Even the MIL rated stuff I've used tops out (free air rated) at typically 100C with ...

+ Reply to Thread
Page 2 of 2 FirstFirst 1 2
Results 21 to 35 of 35

Thread: CPU temp

  1. Re: CPU temp

    Hans-Peter Diettrich wrote:

    > Right, but steam from a wet finger indicates > 100C ;-)


    So does a chip logo branded on your pinky.

    Even the MIL rated stuff I've used tops out (free air rated) at typically
    100C with a few specialty devices going to 125C. Nominal commercial range
    is 0 - 70C.

    OT, but a standing rib roast slow cooks to perfection at 165F for 4-5 hours.

    --
    Will Honea

  2. Re: CPU temp

    Ron Gibson wrote:
    > On Wed, 01 Aug 2007 22:38:16 -0400, Barnacle Bill the Sailor wrote:
    >
    >>> That's a big swing when you consider idle temperature is about 108F
    >>> for me.

    >
    >> I see mine idles steady about 98F. I have the bios set for load
    >> throttling, which slows the clock during idle times, and then cranks it
    >> back up when needed. (and a launcher applet tells me where I am at too,
    >> very handy!) At peak speed today while compiling some new stuff I saw the
    >> temp go to 118F, well it peaked there and in fact bobbed up and down as
    >> the load varied. Glad I got these sensors set up, for it's interesting to
    >> see what happens when.

    >
    > Oh Gkrellm with sensors is a killer app. The CPU is conjunction with
    > processes, memory and what top says can help you quickly spot problem
    > apps, ones hogging the CPU or with a memory leak.
    >
    > I like the ETH0 bandwidth monitor as I can set it up to be downloading a
    > torrent with Azureus so as to give the download 150KB/s and still have
    > enough left over to play UT2004. I have a 3MB connection so I have
    > enough to bandwidth to do that with needing a more expensive 10 or 15MB
    > connection.
    >
    > IMO, this is really a good thing to have the info to tweak stuff so you
    > get the maximum out of your machine.
    >
    > Here's a wild thought I had. I'm an engineer which means I have
    > terminal OCD The way that /etc/sensors.conf is set up you could
    > take exact direct readings with a thermometer and a voltage meter and
    > REALLY fine tune it all, LOL!
    >
    > I try to make a point of informing people about this stuff as
    > surprisingly enough a lot of people aren't aware of these capabilities.
    >
    > Enjoy...
    >

    Hmmmmmmm....a wee little temp probe.....think think think! Not an
    engineer but I missed my calling. Well, yes, I am... a Locomotive Engineer.

  3. Re: CPU temp

    Barnacle Bill the Sailor writes:

    >> get the maximum out of your machine.
    >>
    >> Here's a wild thought I had. I'm an engineer which means I have
    >> terminal OCD The way that /etc/sensors.conf is set up you could
    >> take exact direct readings with a thermometer and a voltage meter and
    >> REALLY fine tune it all, LOL!


    The temp probes are inside the chip , not on the surface. Thus the
    conduction of the package becomes important.

    >>
    >> I try to make a point of informing people about this stuff as
    >> surprisingly enough a lot of people aren't aware of these capabilities.
    >>
    >> Enjoy...
    >>

    >Hmmmmmmm....a wee little temp probe.....think think think! Not an
    >engineer but I missed my calling. Well, yes, I am... a Locomotive Engineer.


  4. Re: CPU temp

    On Thu, 02 Aug 2007 12:50:38 -0600, Will Honea wrote:

    > OT, but a standing rib roast slow cooks to perfection at 165F for 4-5
    > hours.


    I'm a hardcore bachelor. I got some good recipies for us lazy dudes too
    :-)

    --
    Linux Help: http://rsgibson.com/linux.htm
    Email - rsgibson@verizon.borg
    Replace borg with net


  5. Re: CPU temp

    On Thu, 02 Aug 2007 21:04:56 +0000, Unruh wrote:

    >>> Here's a wild thought I had. I'm an engineer which means I have
    >>> terminal OCD The way that /etc/sensors.conf is set up you could
    >>> take exact direct readings with a thermometer and a voltage meter and
    >>> REALLY fine tune it all, LOL!


    > The temp probes are inside the chip , not on the surface. Thus the
    > conduction of the package becomes important.


    Like I said you can make this as complicated as you like. Since the
    thermal properties of the material are well known (or if not could be
    empirically determined) the internal temperature could be calculated from
    the surface temperature with a high degree of precision. I've done those
    types of calculations many times.

    But lets don't confuse our original poster. For using lm_sensors
    calibrating the sensor readings in Linux with BIOS readings is plenty
    good enough

    --
    Linux Help: http://rsgibson.com/linux.htm
    Email - rsgibson@verizon.borg
    Replace borg with net


  6. Re: CPU temp

    Ron Gibson writes:

    >On Thu, 02 Aug 2007 21:04:56 +0000, Unruh wrote:


    >>>> Here's a wild thought I had. I'm an engineer which means I have
    >>>> terminal OCD The way that /etc/sensors.conf is set up you could
    >>>> take exact direct readings with a thermometer and a voltage meter and
    >>>> REALLY fine tune it all, LOL!

    >
    >> The temp probes are inside the chip , not on the surface. Thus the
    >> conduction of the package becomes important.

    >
    >Like I said you can make this as complicated as you like. Since the
    >thermal properties of the material are well known (or if not could be
    >empirically determined) the internal temperature could be calculated from
    >the surface temperature with a high degree of precision. I've done those
    >types of calculations many times.


    Actually no. Lets take the sun. Lets say that the internal heat production
    doubled and the internal temp shot up by a factor of 10 or so. It would
    take about 1000 years for the surface of the sun to start getting hotter.
    If that were your chip it would have fried long before you noticed anything
    on the case. Ie, there is a time lag between the internal temp and the temp
    on the outside of the case, and there is no way to compensate for that.


    >But lets don't confuse our original poster. For using lm_sensors
    >calibrating the sensor readings in Linux with BIOS readings is plenty
    >good enough


    >--
    >Linux Help: http://rsgibson.com/linux.htm
    >Email - rsgibson@verizon.borg
    >Replace borg with net



  7. Re: CPU temp

    On Fri, 03 Aug 2007 02:00:33 +0000, Unruh wrote:

    >>Like I said you can make this as complicated as you like. Since the
    >>thermal properties of the material are well known (or if not could be
    >>empirically determined) the internal temperature could be calculated from
    >>the surface temperature with a high degree of precision. I've done those
    >>types of calculations many times.


    > Actually no. Lets take the sun. Lets say that the internal heat production
    > doubled and the internal temp shot up by a factor of 10 or so. It would
    > take about 1000 years for the surface of the sun to start getting hotter.
    > If that were your chip it would have fried long before you noticed
    > anything on the case. Ie, there is a time lag between the internal temp
    > and the temp on the outside of the case, and there is no way to compensate
    > for that.


    Listen, I'm a mechanical engineer and heat transfer and thermodynamics
    is our speciality. I can tell by your post you don't understand heat
    transfer so lets drop this.

    --
    Linux Help: http://rsgibson.com/linux.htm
    Email - rsgibson@verizon.borg
    Replace borg with net


  8. Re: CPU temp

    In Ron Gibson:

    [Snip...]

    > Listen, I'm a mechanical engineer and heat transfer and thermodynamics
    > is our speciality. I can tell by your post you don't understand heat
    > transfer so lets drop this.


    FWIW...

    http://en.wikipedia.org/wiki/Bill_Unruh

    YMMV; HTH; HAND...

    --
    Regards, Weird (Harold Stevens) * IMPORTANT EMAIL INFO FOLLOWS *
    Pardon any bogus email addresses (wookie) in place for spambots.
    Really, it's (wyrd) at airmail, dotted with net. DO NOT SPAM IT.
    Kids jumping ship? Looking to hire an old-school type? Email me.

  9. Re: CPU temp

    Ron Gibson wrote:

    >>The temp probes are inside the chip , not on the surface. Thus the
    >>conduction of the package becomes important.

    >
    >
    > Like I said you can make this as complicated as you like. Since the
    > thermal properties of the material are well known (or if not could be
    > empirically determined) the internal temperature could be calculated from
    > the surface temperature with a high degree of precision. I've done those
    > types of calculations many times.


    Surface measurments cannot take into account hot spots on the chip,
    which can result in a local damage, even if the overall temperature is
    within the spacified range.

    As a practical example: with a welding torch you can burn a hole into
    some material, while holding it on another edge without burning your
    fingers.

    > But lets don't confuse our original poster. For using lm_sensors
    > calibrating the sensor readings in Linux with BIOS readings is plenty
    > good enough


    The readings will mostly be good enough, for display purposes ;-)

    DoDi

  10. Re: CPU temp

    Ron Gibson writes:

    >On Fri, 03 Aug 2007 02:00:33 +0000, Unruh wrote:
    >
    >>>Like I said you can make this as complicated as you like. Since the
    >>>thermal properties of the material are well known (or if not could be
    >>>empirically determined) the internal temperature could be calculated from
    >>>the surface temperature with a high degree of precision. I've done those
    >>>types of calculations many times.

    >
    >> Actually no. Lets take the sun. Lets say that the internal heat production
    >> doubled and the internal temp shot up by a factor of 10 or so. It would
    >> take about 1000 years for the surface of the sun to start getting hotter.
    >> If that were your chip it would have fried long before you noticed
    >> anything on the case. Ie, there is a time lag between the internal temp
    >> and the temp on the outside of the case, and there is no way to compensate
    >> for that.

    >
    >Listen, I'm a mechanical engineer and heat transfer and thermodynamics
    >is our speciality. I can tell by your post you don't understand heat
    >transfer so lets drop this.


    OOO. The old "appeal to authority" in this case his own. In steady state
    there is a close relation between on chip temp and case temp. In dynamic situations there
    is not. There is a lag and a smearing. If you do not understand that then you are a bust as a mechanical
    engineer and I pity your clients.



  11. Re: CPU temp

    On Fri, 03 Aug 2007 18:11:50 +0000, Unruh wrote:

    >>Listen, I'm a mechanical engineer and heat transfer and thermodynamics
    >> is
    >>our specialty. I can tell by your post you don't understand heat
    >>transfer so lets drop this.


    > OOO. The old "appeal to authority" in this case his own. In steady state
    > there is a close relation between on chip temp and case temp. In dynamic
    > situations there is not. There is a lag and a smearing. If you do not
    > understand that then you are a bust as a mechanical engineer and I pity
    > your clients.


    The lag time is trivial and can be calculated. That "close relation" is
    known as a transfer function.

    Thanx but no cigar.

    Bye now.

    --
    Linux Help: http://rsgibson.com/linux.htm
    Email - rsgibson@verizon.borg
    Replace borg with net


  12. Re: CPU temp

    Ron Gibson wrote:
    > On Fri, 03 Aug 2007 02:00:33 +0000, Unruh wrote:
    >
    >>> Like I said you can make this as complicated as you like. Since the
    >>> thermal properties of the material are well known (or if not could be
    >>> empirically determined) the internal temperature could be calculated from
    >>> the surface temperature with a high degree of precision. I've done those
    >>> types of calculations many times.

    >
    >> Actually no. Lets take the sun. Lets say that the internal heat production
    >> doubled and the internal temp shot up by a factor of 10 or so. It would
    >> take about 1000 years for the surface of the sun to start getting hotter.
    >> If that were your chip it would have fried long before you noticed
    >> anything on the case. Ie, there is a time lag between the internal temp
    >> and the temp on the outside of the case, and there is no way to compensate
    >> for that.

    >
    > Listen, I'm a mechanical engineer and heat transfer and thermodynamics
    > is our speciality. I can tell by your post you don't understand heat
    > transfer so lets drop this.
    >



    And I am plucky enough to dig around in there with a decent probe just
    to see what I get.

  13. Re: CPU temp

    Ron Gibson wrote:
    > On Fri, 03 Aug 2007 18:11:50 +0000, Unruh wrote:
    >
    >>> Listen, I'm a mechanical engineer and heat transfer and thermodynamics
    >>> is
    >>> our specialty. I can tell by your post you don't understand heat
    >>> transfer so lets drop this.

    >
    >> OOO. The old "appeal to authority" in this case his own. In steady state
    >> there is a close relation between on chip temp and case temp. In dynamic
    >> situations there is not. There is a lag and a smearing. If you do not
    >> understand that then you are a bust as a mechanical engineer and I pity
    >> your clients.

    >
    > The lag time is trivial and can be calculated. That "close relation" is
    > known as a transfer function.
    >
    > Thanx but no cigar.
    >
    > Bye now.
    >

    OH such a monster I've created....now let's see...there's things in my
    box which can heat up. There are sensors in said things which give me
    some number. Firstly, how to fetch those numbers. OK got that...results
    are subject to adjustment. And curiosity which is said to have maimed
    felines, makes me want to dig around with a temp probe and just see what
    I can see. Though in the end esteemed gentlemen...With whatever info I
    can get I must make a leap of faith and declare it normal. Now all I
    have to do is monitor it to see that it doesn't rise willy nilly. If I
    want to find the true temperature of various bits, I can accept the
    declrations of the BIOS or try on my own, but I only HAVE to decide how
    "normal" shall be regarded.

  14. Re: CPU temp

    On Sat, 04 Aug 2007 14:37:25 -0400, Barnacle Bill the Sailor wrote:

    > Though in the end esteemed gentlemen...With whatever info I can get I
    > must
    > make a leap of faith and declare it normal. Now all I have to do is
    > monitor it to see that it doesn't rise willy nilly. If I want to find the
    > true temperature of various bits, I can accept the declrations of the BIOS
    > or try on my own, but I only HAVE to decide how "normal" shall be
    > regarded.


    Just go by the BIOS settings as the true settings. I just threw that out
    there because it was interesting to me having spent time doing a lot of
    that stuff.

    Engineers make assumptions and accept a reasonable degree of accuracy.
    We take a macroscopic view. I'm not interested in what the temperature is
    10 microns apart at different points. A single sensor could not measure
    that anyway.

    --
    Linux Help: http://rsgibson.com/linux.htm
    Email - rsgibson@verizon.borg
    Replace borg with net


  15. Re: CPU temp

    On Sat, 04 Aug 2007 11:13:56 -0400, Barnacle Bill the Sailor wrote:

    > And I am plucky enough to dig around in there with a decent probe just
    > to see what I get.


    It would be kind of fun :-)

    --
    Linux Help: http://rsgibson.com/linux.htm
    Email - rsgibson@verizon.borg
    Replace borg with net


+ Reply to Thread
Page 2 of 2 FirstFirst 1 2