<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>Heat fins are used to increase the surface area used for heat
      transfer, since the rate of energy transfer by conduction is
      directly proportional the surface area. Heat fins are needed when
      air is involved because air has such a low thermal conductivity. <br>
    </p>
    <p>Thermal conductivity of liquids are much high, so heat fins
      aren't as necessary. For example, I've read that water can
      transfer heat orders of magnitude better than air, so using water
      to remove hear from a processor would need orders of magnitude
      less surface area for the same energy transfer rate. <br>
    </p>
    <p>Also, liquids have higher viscosities than gases, so we have to
      worry about 'boundary layers'. A boundary layer is area where the
      edge flowing fluid is in contact with a solid. The friction
      between the liquid and the solid slows down the fluid near the
      solid. This affects both gases and liquids, but since liquids have
      higher viscosities, the effect is more noticeable. <br>
    </p>
    <p>Think about a car's radiator - the air side has all the fins on
      it, and the liquid side has smooth pipe walls. <br>
    </p>
    <p><a class="moz-txt-link-freetext" href="https://en.wikipedia.org/wiki/Boundary_layer">https://en.wikipedia.org/wiki/Boundary_layer</a></p>
    <p>Convection heat transfer is an equally important mode of heat
      transfer in fluids, and in the boundary layer, where the liquids
      aren't moving as fast, heat transfer isn't as good, so you need to
      keep your boundary layer from becoming too thick. <br>
    </p>
    <p>Since fluids have much higher thermal conductivities, and
      boundary layer effects are more of a concern, I actually think a
      smooth heat transfer surface would be better in these immersion
      cooling cases. I'm sure smaller,  more spaced out fins would
      probably help heat transfer without  creating too much of a
      boundary layer, but making those heat sinks adds cost for
      increased performance in a situation where it probably isn't
      needed. <br>
    </p>
    <p>Now direct-contact cooling systems like Asetek products  do have
      fins on the liquid side, if I remember correctly, but that in
      those systems, there are pumps to provide forced convection. In
      immersion cooling, you are relying on natural convection, so there
      isn't as much driving force to overcome viscosity/boundary layer
      effects to force the liquid through the heat fins. <br>
    </p>
    <p>That's my thoughts, anyway. <br>
    </p>
    <pre class="moz-signature" cols="72">Prentice </pre>
    <div class="moz-cite-prefix">On 11/07/2018 04:12 AM, John Hearns via
      Beowulf wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAPqNE2XO_99NQ-3GP1ZixwfFoFzmFUGNMKErDHdRxHWB8e=AzQ@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=utf-8">
      <div dir="ltr">
        <div dir="ltr">
          <div>Thinking about liquid cooling , and the ebuillient
            cooling, the main sources of heat on our current
            architecture servers are the CPU package and the voltage
            regulators. Then the DIMMs.</div>
          <div>Concentrating on the CPU die package, it is engineered
            with a flat metal surface which is intended to have a
            thermal paste to transfer heat across to a flat metal
            heatsink.</div>
          <div>Those heatsinks are finned to have air blown across them
            to transport the heat away.</div>
          <div><br>
          </div>
          <div>In liquid immersion should we be looking at having a
            spiky surface on the CPU die packages and the voltage
            regulators?</div>
          <div>Maybe we should spray the entire board with a 'flocking''
            compound and give it a matt finish!</div>
          <div>I am being semi-serious. I guess a lot of CFD simulation 
            done regarding air cooling with fins.</div>
          <div>How much work has gone into pointy surfaces on the die
            package, which would increase contact area of course and
            also act as nucleation points for bubbles?</div>
          <div><br>
          </div>
          <div>One interesting experiment to do - assuming the flat
            areas of the CPU in an immersive system do not have (non
            thermal paste) heatsinks bolted on:</div>
          <div>take two systems and roughen up the die package
            surfacewith sandpaper on one. Compare temperatures.</div>
          <div><br>
          </div>
          <div>ps. I can't resist adding this. Sorry Stu .   <a
              href="https://www.youtube.com/watch?v=kHnifVTSFEo"
              moz-do-not-send="true">https://www.youtube.com/watch?v=kHnifVTSFEo</a></div>
          <div>I guess Kenneth Williams is a typical vendor Site
            Engineer.</div>
          <div>pps. the actress in the redress had her career ruined by
            this film - she ver got a serious role again after perfectly
            being typecast.</div>
          <div><br>
          </div>
          <div><br>
          </div>
          <div><br>
          </div>
          <div><br>
          </div>
          <div><br>
          </div>
          <div><br>
          </div>
        </div>
      </div>
      <br>
      <div class="gmail_quote">
        <div dir="ltr">On Tue, 6 Nov 2018 at 22:57, Prentice Bisbal via
          Beowulf <<a href="mailto:beowulf@beowulf.org"
            moz-do-not-send="true">beowulf@beowulf.org</a>> wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0 0 0
          .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <div text="#000000" bgcolor="#FFFFFF">
            <div class="m_4669408001860479086moz-cite-prefix">On
              11/06/2018 02:03 PM, Lux, Jim (337K) wrote:<br>
            </div>
            <blockquote type="cite">
              <div class="m_4669408001860479086WordSection1">
                <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d">True
                    enough.</span></p>
                <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d">Ebullient
                    cooling does have some challenges – you can form
                    vapor films, which are good insulators, but if you
                    get the system working right, nothing beats phase
                    changes for a heat transfer.  </span></p>
              </div>
            </blockquote>
            If I recall what I learned in my Transport Phenomena classes
            in engineering school, you need a reasonably high
            temperature difference to get a stable film like that. For
            that to happen, radiant heat transfer needs to be the
            dominant heat transfer mechanism, in the range of operation
            we are talking about, the temperature difference isn't that
            great, and conduction is still the dominant form of heat
            transfer. <br>
            <br>
            Here's an example of what 3M Novec ebullient cooling looks
            like. It doesn't look like it's anywhere near the film
            boiling regime: <br>
            <br>
            <a class="m_4669408001860479086moz-txt-link-freetext"
              href="https://www.youtube.com/watch?v=CIbnl3Pj15w"
              target="_blank" moz-do-not-send="true">https://www.youtube.com/watch?v=CIbnl3Pj15w</a><br>
            <br>
            --<br>
            Prentice<br>
            <br>
            <blockquote type="cite">
              <div class="m_4669408001860479086WordSection1"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"></span>
                <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"> </span></p>
                <span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"></span>
                <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"> </span></p>
                <div>
                  <div style="border:none;border-top:solid #e1e1e1
                    1.0pt;padding:3.0pt 0in 0in 0in">
                    <p class="MsoNormal"><b><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:windowtext">From:</span></b><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:windowtext">
                        Beowulf [<a
                          class="m_4669408001860479086moz-txt-link-freetext"
                          href="mailto:beowulf-bounces@beowulf.org"
                          target="_blank" moz-do-not-send="true">mailto:beowulf-bounces@beowulf.org</a>]
                        <b>On Behalf Of </b>Prentice Bisbal via Beowulf<br>
                        <b>Sent:</b> Tuesday, November 06, 2018 8:17 AM<br>
                        <b>To:</b> <a
                          class="m_4669408001860479086moz-txt-link-abbreviated"
                          href="mailto:beowulf@beowulf.org"
                          target="_blank" moz-do-not-send="true">beowulf@beowulf.org</a><br>
                        <b>Subject:</b> Re: [Beowulf] More about those
                        underwater data centers</span></p>
                  </div>
                </div>
                <p class="MsoNormal"> </p>
                <blockquote style="margin-top:5.0pt;margin-bottom:5.0pt">
                  <p class="MsoNormal">. And serviceability is
                    challenging. You need to pull the "wet" boards out,
                    or you need to connect and disconnect fluid
                    connectors, etc.  If you're in an environment where
                    you can manage that (or are forced into it by
                    necessity), then you can do it.</p>
                </blockquote>
                <p>I think everyone on this list already knows I'm no
                  fan of mineral oil immersion (It just seems to messy
                  to me. Sorry, Stu), but immersion cooling with other
                  liquids, such as 3M Novec engineered fluid addresses a
                  lot of your concerns. It as a low boiling point, not
                  much above room temperature, and it was originally
                  meant to be an electronic parts cleaner (according to
                  a 3M rep at the 3M booth at SC a few years ago, so if
                  you pull a component out of it, it dries very quickly
                  and should be immaculately clean. </p>
                <p>The low boiling point is an excellent feature for
                  heat transfer, too, since it boils from the heat of
                  the processor (ebullient cooling). This change of
                  state absorbs a lot of energy, making it very
                  effective at transferring heat away from the
                  processor. The vapor can then rise and condense on a
                  heat exchanger with a chilled water heat exchanger,
                  where it again transfers a lot of heat through a
                  change of state. </p>
                <pre>Prentice </pre>
                <div>
                  <p class="MsoNormal">On 11/05/2018 06:30 PM, Stu
                    Midgley wrote:</p>
                </div>
                <blockquote style="margin-top:5.0pt;margin-bottom:5.0pt">
                  <div>
                    <div>
                      <p class="MsoNormal">I refute both these claims.</p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal">You DO want to run your
                        boards immersed in coolant.  It works
                        wonderfully well, is easy to live with,
                        servicing is easy... and saves you almost 1/2
                        your power bill.</p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal">People are scared of
                        immersion cooling, but it isn't that difficult
                        to live with.  Some things are harder but other
                        things are way easier.  In total, it balances
                        out.</p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal">Also, given the greater
                        reliability of components you get, you do less
                        servicing.</p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal">If you haven't lived with it,
                        you really have no idea what you are missing.</p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal">Serviceability is NOT
                        challenging.</p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <div>
                      <p class="MsoNormal"> </p>
                    </div>
                    <p class="MsoNormal"> </p>
                    <div>
                      <blockquote style="border:none;border-left:solid
                        #cccccc 1.0pt;padding:0in 0in 0in
                        6.0pt;margin-left:4.8pt;margin-right:0in">
                        <p class="MsoNormal">You really do NOT want to
                          run boards immersed in coolant - yeah, there's
                          folks doing it at HPC scale<br>
                          <br>
                          Whatever the coolant, it leaks, it oozes, it
                          gets places you don't want it to go. And
                          serviceability is challenging. You need to
                          pull the "wet" boards out, or you need to
                          connect and disconnect fluid connectors, etc. 
                          If you're in an environment where you can
                          manage that (or are forced into it by
                          necessity), then you can do it.</p>
                      </blockquote>
                    </div>
                    <p class="MsoNormal">-- </p>
                    <div>
                      <div>
                        <p class="MsoNormal">Dr Stuart Midgley<br>
                          <a href="mailto:sdm900@gmail.com"
                            target="_blank" moz-do-not-send="true">sdm900@gmail.com</a></p>
                      </div>
                    </div>
                  </div>
                  <p class="MsoNormal"><br>
                    <br>
                    <br>
                  </p>
                  <pre>_______________________________________________</pre>
                  <pre>Beowulf mailing list, <a href="mailto:Beowulf@beowulf.org" target="_blank" moz-do-not-send="true">Beowulf@beowulf.org</a> sponsored by Penguin Computing</pre>
                  <pre>To change your subscription (digest mode or unsubscribe) visit <a href="http://www.beowulf.org/mailman/listinfo/beowulf" target="_blank" moz-do-not-send="true">http://www.beowulf.org/mailman/listinfo/beowulf</a></pre>
                </blockquote>
                <p class="MsoNormal"> </p>
              </div>
            </blockquote>
            <br>
          </div>
          _______________________________________________<br>
          Beowulf mailing list, <a href="mailto:Beowulf@beowulf.org"
            target="_blank" moz-do-not-send="true">Beowulf@beowulf.org</a>
          sponsored by Penguin Computing<br>
          To change your subscription (digest mode or unsubscribe) visit
          <a href="http://www.beowulf.org/mailman/listinfo/beowulf"
            rel="noreferrer" target="_blank" moz-do-not-send="true">http://www.beowulf.org/mailman/listinfo/beowulf</a><br>
        </blockquote>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
Beowulf mailing list, <a class="moz-txt-link-abbreviated" href="mailto:Beowulf@beowulf.org">Beowulf@beowulf.org</a> sponsored by Penguin Computing
To change your subscription (digest mode or unsubscribe) visit <a class="moz-txt-link-freetext" href="http://www.beowulf.org/mailman/listinfo/beowulf">http://www.beowulf.org/mailman/listinfo/beowulf</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>