1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

Interesting tidbit

Discussion in 'AMD Thunderbird' started by Bill, Jul 13, 2003.

  1. Bill

    Bill Guest


    Quoted From the "AMD Athlon™ 64 Processor Thermal Design Guide"

    Page 22 Thermal Interface Material

    "The heat sink makes contact with the top surface of the processor
    package utilizing the thermal interface material between the processor
    lid and the heat sink. AMD recommends using a high-performance grease
    such as those listed in Table 6. AMD does not recommend using phase-
    change materials between the heat sink and the processor.
    Phase-change materials develop high adhesion forces between the heat
    sink and processor when the material is in the solid phase. This
    strong adhesive force may cause the processor to stick to the heat
    sink. During heat sink removal, this strong adhesive force may
    cause the processor to be removed from the socket while it
    is locked, and this action can result in damage to the socket or to the
    processor pins."

    Bill, Jul 13, 2003
    1. Advertisements

  2. Interesting as I think there was a thread in one of the newsgroups several
    months ago discussing almost the opposite recommendation about thermal
    paste. That was based on an AMD statement in regards to creepage(?) of
    thermal paste due to temperature changes over a period of time.

    I can only guess that several people have made a warranty claim for their
    CPUs based on damage related to thermal pads.

    Also interesting is I bought a new XP 2500+ retail packaged processor about
    2 weeks ago and the supplied heatsink came with a thermal pad already
    applied to it!!!!
    callsignviper, Jul 13, 2003
    1. Advertisements

  3. Bill

    Bill Guest

    Yep, I remember that. Now AMD goes 180 out on the subject.
    Luckily, they come off fairly easily.

    Bill, Jul 14, 2003
  4. Bill

    Ed Light Guest

    My first heatsink install involved getting a thermal pad off the cpu which
    was much like epoxy! I actually had to scrape it. It didn't want to soften
    up. It came on a Speeze sink. I was really relieved when that cpu fired up

    So, I'm a greaser forever.

    Besides, online test results show that after Arctic whatever breaks in it
    gives 2 or 3 degrees better cooling.

    Arctic Alumina's supposed to come off easier than Silver, and it's just $4.
    Ed Light, Jul 14, 2003
  5. Bill

    Ed Light Guest

    I might have said that this thermal pad had been on there for a few months.
    The Duron was running in the 40's, as I remember. It became epoxy.
    Ed Light, Jul 14, 2003
  6. Bill

    Stacey Guest

    Funny to me how a company as big as AMD can publish BS like that! At one
    point they even said they wouldn't warranty a chip that had thermal grease
    used on it?
    Stacey, Jul 15, 2003
  7. Bill

    J.Clarke Guest

    The purpose of the grease is to provide something better than air
    connecting those places where solid-to-solid contact does not occur due
    to surface defects and roughness. Too much grease and there is no
    solid-to-solid contact, everything goes through the grease. And the
    grease, though a lot better than air, is not nearly as good as direct
    J.Clarke, Jul 15, 2003
  8. Bill

    Soupladel Guest

    Such as?

    Just out of interest

    Soupladel, Jul 15, 2003
  9. Bill

    J.Clarke Guest

    I believe that they have changed the design a bit--the Athlons have
    exposed cores while I believe the Opterons have a heat spreader. The
    spreader provides greater surface area and so increases the strength of
    the bond achieved by the phase-change material and so might cause the
    problems they describe. With the Athlons the problem is
    different--there is limited contact surface so the highest-possible
    quality of contact is desirable.

    The notion that grease "creeps" is certainly accurate. Glass creeps if
    you give it enough time, and thermal grease is a lot weaker than glass.
    The real question is whether it is going to creep quickly enough to
    cause problems, and the answer is (in both cases) no--for glass we're
    talking geological time, for grease a lot longer than the processor is
    likely to be useful.
    J.Clarke, Jul 16, 2003
  10. Bill

    Stacey Guest

    Then why have they turned around and are now suggesting NOT to use a thermal
    pad? And do you believe thermal grease "creeps"?
    Stacey, Jul 16, 2003
  11. Bill

    Ben Pope Guest

    Quality thermal paste is fine, but there is some real crap out there.

    Thermal pads aren't that great and the original Opterons will need to
    dissipate something like 90W, I think.

    As J.Clarke has said, the design is different, heat spreders on the chip
    will help to reduce problems associated with clumsy application of thermal

    Ben Pope, Jul 16, 2003
  12. Bill

    Stacey Guest

    J.Clarke wrote:

    So you believe that urban legend?


    "The calculation showed that if a plate of glass a meter tall and a
    centimeter thick was placed in an upright position at room temperature, the
    time required for the glass to flow down so as to thicken 10 angstrom units
    at the bottom (a change the size of only a few atoms) would theoretically
    be about the same as the age of the universe: close to ten billion years."

    Exactly, the creep thing posted by AMD is BS and I find it sad they posted
    info like that as fact.
    Stacey, Jul 17, 2003
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.