<div dir="ltr"><div dir="ltr"><div dir="ltr"><div>Prentice, the website refers to Open Compute racks.  "... technology has been designed to fit into standard Open Compute racks".</div><div>So yep, 19 inch racks are not being targeted here. But OCP is pretty widespread.</div><div>I would really like to find out if they can retrofit these to existing kit. I suspect though that you need servers engineered to fit onto their heatsinks.</div><div><span style="text-align:left;color:rgb(255,255,255);text-transform:none;text-indent:0px;letter-spacing:0px;font-family:myriad-pro;font-size:22px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;word-wrap:break-word;float:none;background-color:transparent"><span style="text-align:left;color:rgb(255,255,255);text-transform:none;text-indent:0px;letter-spacing:0px;font-family:myriad-pro;font-size:22px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;word-wrap:break-word;float:none;background-color:transparent">orced Physics cooling technology has been designed to fit into Standard Open Compute Racks. </span>orced Physics cooling technology has been designed to fit into Standard Open Compute Racks. </span></div><div><b></b><i></i><u></u><sub></sub><sup></sup><strike></strike><br></div></div></div></div><br><div class="gmail_quote"><div class="gmail_attr" dir="ltr">On Mon, 28 Jan 2019 at 16:48, Prentice Bisbal via Beowulf <<a href="mailto:beowulf@beowulf.org">beowulf@beowulf.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;padding-left:1ex;border-left-color:rgb(204,204,204);border-left-width:1px;border-left-style:solid">
  
    
  
  <div bgcolor="#FFFFFF">
    <p>So I was thinking about this over the weekend (because I
      apparently have nothing better to do with my time), and I
      definitely think this is a non-starter due to the massive change
      in server hardware layout to accommodate this thing. Yes, blades,
      and twin form factor servers already required that, and they're
      common form factors, but those form factors were just a matter of
      shrinking or changing the layout of the motherboard but still look
      like "traditional" layouts to the untrained eye, and they were
      still designed with typical front-to-back air cooling in mind. I
      feel like re-arranging the layout of components to accomodate this
      thing is a just more of a change than the market will accept. <br>
    </p>
    <p>Just my 2 cents. <br>
    </p>
    <pre class="gmail-m_-4339801130891775252moz-signature" cols="72">Prentice </pre>
    <div class="gmail-m_-4339801130891775252moz-cite-prefix">On 1/25/19 3:56 PM, Prentice Bisbal
      wrote:<br>
    </div>
    <blockquote type="cite">
      
      <p>Eric, <br>
      </p>
      <p>I was suspecting that might be the case, but the explanations
        in the other articles were way too vague to be sure of that. The
        NextPlatform provided much better pictures. If that's the case,
        this thing operates like a direct-expansion (DX) refrigeration
        system, where the refrigerant is air and does not change state
        from liquid to gas, like a typical DX refrigeration system, and
        the induced-draft fan provides the shaft work, and those tiny
        channels that allegedly line up the molecules act as many tiny
        offices for the throttling process. Based on the pictures in the
        Next Platform article, here is a crude drawing of cross-section
        of one of these devices that I drew in Google Draw. It should
        help you understand what's going inside this thing: <br>
      </p>
      <p><a class="gmail-m_-4339801130891775252moz-txt-link-freetext" href="https://docs.google.com/document/d/1UK94PxVlQtVSb2ns5TbCqHjPJ1vYSOmkGSeSorvHyaM/edit?usp=sharing" target="_blank">https://docs.google.com/document/d/1UK94PxVlQtVSb2ns5TbCqHjPJ1vYSOmkGSeSorvHyaM/edit?usp=sharing</a></p>
      <p>Given this design, you can only have an induced-draft fan on
        the outlet. A forced-draft fan on the inlet would compress the
        air, heating it up and negating the throttling (or
        Joule-Thompson) effect on the low-pressure side. <br>
      </p>
      <p>At the end of the day, thermodynamics still says X amount of
        shaft work has to be done to provide Y amount of cooling through
        this process, so I'm still skeptical of it, especially at scale.
        <br>
      </p>
      <p>And for those of you looking for something really boring to
        read rather than work, here are the related patents. I haven't
        read them myself. <br>
      </p>
      <p><a class="gmail-m_-4339801130891775252moz-txt-link-freetext" href="https://patents.google.com/patent/US8414847" target="_blank">https://patents.google.com/patent/US8414847</a></p>
      <p><a class="gmail-m_-4339801130891775252moz-txt-link-freetext" href="https://patents.google.com/patent/US8986627B2" target="_blank">https://patents.google.com/patent/US8986627B2</a></p>
      <p><a class="gmail-m_-4339801130891775252moz-txt-link-freetext" href="https://patents.google.com/patent/US10113774B2" target="_blank">https://patents.google.com/patent/US10113774B2</a><br>
      </p>
      <pre class="gmail-m_-4339801130891775252moz-signature" cols="72">Prentice</pre>
      <div class="gmail-m_-4339801130891775252moz-cite-prefix">On 1/25/19 2:26 PM, Eric Moore wrote:<br>
      </div>
      <blockquote type="cite">
        
        <div dir="ltr">
          <div dir="ltr">Actually, it looks like Joule-Thompson cooling
            to me (Especially given the "Joule Force" name). You've got
            the air intake (ambient), then an expansion nozzle, into a
            low-pressure region, which is created by the fan at the end.
            So the outlet velocity of the air (and thus it's kinetic
            energy) is higher than the inlet velocity, which would lower
            the internal energy, and thus the temperature. Instead the
            fins/nozzle/heatsink transfer heat to the expanding gas,
            which exits a little above ambient temperature. I imagine
            the drawback is you really need to get rid of that high
            velocity hot air, and can't recirculate it, or the kinetic
            energy would be converted back to thermal energy, and mess
            it all up. The descriptions do all involve the exhaust air
            being ducted to the outside. This article has the most
            technical detail: <a href="https://www.nextplatform.com/2018/12/04/the-leading-edge-of-air-cooled-servers-leads-to-the-edge/" target="_blank">https://www.nextplatform.com/2018/12/04/the-leading-edge-of-air-cooled-servers-leads-to-the-edge/</a></div>
        </div>
        <br>
        <div class="gmail_quote">
          <div class="gmail-m_-4339801130891775252gmail_attr" dir="ltr">On Fri, Jan 25, 2019 at
            11:33 AM Prentice Bisbal via Beowulf <<a href="mailto:beowulf@beowulf.org" target="_blank">beowulf@beowulf.org</a>>
            wrote:<br>
          </div>
          <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;padding-left:1ex;border-left-color:rgb(204,204,204);border-left-width:1px;border-left-style:solid">
            <div bgcolor="#FFFFFF">
              <p>You all know how much I like talking about heat
                transfer and server cooling, so I decided to do some
                research on this product:</p>
              <p>Here's their website: <br>
              </p>
              <p><a class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-txt-link-freetext" href="https://forcedphysics.com" target="_blank">https://forcedphysics.com</a><br>
              </p>
              <p>and here's their YouTube channel with 5 videos:<br>
              </p>
              <p><a class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-txt-link-freetext" href="https://www.youtube.com/channel/UClwWeahYGuNl0THWVz1Hyow/videos" target="_blank">https://www.youtube.com/channel/UClwWeahYGuNl0THWVz1Hyow/videos</a> </p>
              <p>This is really nothing more than an air-cooled
                heatsink. I'm afraid I'm going to have to call BS on
                this technology for the following reasons: <br>
              </p>
              <p>1. It still uses air as the primary cooling medium. I
                just don't think air has adequate thermal conductivity
                or thermal capacity to serve modern processor, no matter
                what you do to it. <br>
              </p>
              <p>2. In the videos, they present highly idealized tests
                with no control to use for comparison. How do I know I
                wouldn't get the same results doing the same experiment
                but using a similar duct fashioned out of sheet metal. <br>
              </p>
              <p>3. Using this technology means a complete redesign of
                your server hardware and possibly your racks.</p>
              <p>4. None of the information in the videos or on their
                website really explains how this technology works, and
                what really differentiates it from any other air-cooled
                heat sink. Most people with a good invention are usually
                excited to tell you how it works. Since they brag about
                30 international patents for this, there's no need to
                try to protect a trade secret. </p>
              <p>5. This statement:</p>
              <p> </p>
              <blockquote type="cite">The fins work like teeth in a
                comb, neatly orienting air molecules to point in the
                same direction and arranging them into columns. </blockquote>
              <p>Based on my education, this statement seems to be
                completely devoid of science. <br>
              </p>
              <p>This statement seems to defy the laws of physics. Last
                time I checked, unless an atom or molecule is at
                absolute zero, it has movement, whether it's spinning or
                vibrating, or both, so how can they get air molecules to
                line up all in neat little rows, where the molecules are
                all pointing the same way? </p>
              <p>This also implies very laminar flow.  As fluid velocity
                increases that the diameter of the channel decreases,
                the Reynolds Number increases. As the Reynold's number
                goes up, turbulence increases, so mathematically, I
                would expect this flow to be tubulent, and not laminar.
                From my classes on heat transfer, turbulent flow around
                the heat transfer surface increases heat transfer, so
                laminar flow in this case wouldn't be a good thing. <br>
              </p>
              <p>Until they can provide better comparisons with real
                servers in real data center environments, I'm going to
                classify this as "snake oil"<br>
              </p>
              <p><a class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-txt-link-freetext" href="https://en.wikipedia.org/wiki/Snake_oil" target="_blank">https://en.wikipedia.org/wiki/Snake_oil</a><br>
              </p>
              <pre class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-signature" cols="72">Prentice</pre>
              <div class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-cite-prefix">On
                1/24/19 3:54 PM, <a class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-txt-link-abbreviated" href="mailto:Chuck_Petras@selinc.com" target="_blank">Chuck_Petras@selinc.com</a>
                wrote:<br>
              </div>
              <blockquote type="cite"> <font face="sans-serif" size="2">Well,
                  this is interesting.</font> <br>
                <br>
                <font face="sans-serif" size="2">"According to Forced
                  Physics’ <</font><a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__forcedphysics.com_&d=DwMFAw&c=-_uRSsrpJskZgEkGwdW-sXvhn_FXVaEGsm0EI46qilk&r=fawF3TRTwCqlaBkoLcxYCr4F4NRwCc64hmEgi9rHPpE&m=zr6lAlVphGxOQTXSElww9hGpqb9IZPik0_MN2v8Fqjs&s=lb4Hi9X8NKIYWe_e1RU3Cw4gr9Uz_B7n5pnCNY0ss3U&e=" target="_blank"><font color="blue" face="sans-serif" size="2">https://forcedphysics.com/</font>
                  [forcedphysics.com]</a><font face="sans-serif" size="2">> chief technology officer, David
                  Binger, the company’s conductor can help a typical
                  data center eliminate its need for water or
                  refrigerants and shrink its 22-MW load by 7.72 MW,
                  which translates to an annual reduction of 67.6
                  million kWh. That data center could also save a total
                  of US $45 million a year on infrastructure, operating,
                  and energy costs with the new system, according to
                  Binger. “We are solving the problem that electrons
                  create,” he said."</font> <br>
                <br>
                <font face="sans-serif" size="2">A Cooler Cloud: A
                  Clever Conduit Cuts Data Centers’ Cooling Needs by 90
                  Percent</font> <br>
                <a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__spectrum.ieee.org_energy_environment_a-2Dcooler-2Dcloud-2Da-2Dclever-2Dconduit-2Dcuts-2Ddata-2Dcenters-2Dcooling-2Dneeds-2Dby-2D90-2Dpercent&d=DwMFAw&c=-_uRSsrpJskZgEkGwdW-sXvhn_FXVaEGsm0EI46qilk&r=fawF3TRTwCqlaBkoLcxYCr4F4NRwCc64hmEgi9rHPpE&m=zr6lAlVphGxOQTXSElww9hGpqb9IZPik0_MN2v8Fqjs&s=VuDTSuinKPMpF6NCztFZkSGOVo3LD7MLjroIj_sn0ao&e=" target="_blank"><font color="blue" face="sans-serif" size="2">https://spectrum.ieee.org/energy/environment/a-cooler-cloud-a-clever-conduit-cuts-data-centers-cooling-needs-by-90-percent</font>
                  [spectrum.ieee.org]</a> <br>
                <font face="sans-serif" size="2"><br>
                  <br>
                  Chuck Petras, PE**<br>
                  Schweitzer Engineering Laboratories, Inc<br>
                  Pullman, WA  99163  USA<br>
                </font><a href="http://www.selinc.com/" target="_blank"><font face="sans-serif" size="2">http://www.selinc.com</font></a><font face="sans-serif" size="2"><br>
                  <br>
                  SEL Synchrophasors - A New View of the Power System
                  <</font><a href="http://synchrophasor.selinc.com/" target="_blank"><font face="sans-serif" size="2">http://synchrophasor.selinc.com</font></a><font face="sans-serif" size="2">><br>
                  <br>
                  Making Electric Power Safer, More Reliable, and More
                  Economical (R)<br>
                  <br>
                  ** Registered in Oregon.<br>
                </font> <br>
                <fieldset class="gmail-m_-4339801130891775252gmail-m_8157281066646160146mimeAttachmentHeader"></fieldset>
                <pre class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-quote-pre">_______________________________________________
Beowulf mailing list, <a class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-txt-link-abbreviated" href="mailto:Beowulf@beowulf.org" target="_blank">Beowulf@beowulf.org</a> sponsored by Penguin Computing
To change your subscription (digest mode or unsubscribe) visit <a class="gmail-m_-4339801130891775252gmail-m_8157281066646160146moz-txt-link-freetext" href="http://www.beowulf.org/mailman/listinfo/beowulf" target="_blank">http://www.beowulf.org/mailman/listinfo/beowulf</a>
</pre>
              </blockquote>
            </div>
            _______________________________________________<br>
            Beowulf mailing list, <a href="mailto:Beowulf@beowulf.org" target="_blank">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" target="_blank" rel="noreferrer">http://www.beowulf.org/mailman/listinfo/beowulf</a><br>
          </blockquote>
        </div>
      </blockquote>
    </blockquote>
  </div>

_______________________________________________<br>
Beowulf mailing list, <a href="mailto:Beowulf@beowulf.org" target="_blank">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" target="_blank" rel="noreferrer">http://www.beowulf.org/mailman/listinfo/beowulf</a><br>
</blockquote></div>