<html>
  <head>
    <meta content="text/html; charset=windows-1252"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <br>
    <div class="moz-cite-prefix">On 09/22/2014 08:16 PM, Khalid
      Baheyeldin wrote:<br>
    </div>
    <blockquote
cite="mid:CA+TuoW1nB67tX3mEVg7e1s0OZ-6KbQPisu5rBH6SHJhVbfX_hw@mail.gmail.com"
      type="cite">
      <div dir="ltr">On Mon, Sep 22, 2014 at 7:59 PM, William Park <span
          dir="ltr"><<a moz-do-not-send="true"
            href="mailto:opengeometry@yahoo.ca" target="_blank">opengeometry@yahoo.ca</a>></span>
        wrote:<br>
        <div class="gmail_extra">
          <div class="gmail_quote">
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex"><span class="">
              </span>- What kind of test?<br>
            </blockquote>
            <div><br>
            </div>
            <div>Copying around 100GB worth of files, most of them
              multi-gigabyte in size, from one disk to the other. Using
              rsync, while the files do not exist on the target<br>
              <br>
            </div>
            <div>rsync -av /mnt/disk1/ /mnt/disk2 <br>
            </div>
            <div dir="rtl"> <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">
              - Sequential, random?<br>
            </blockquote>
            <div><br>
            </div>
            <div>Sequential.<br>
               <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">
              - DC or USB powered?<br>
            </blockquote>
            <div><br>
            </div>
            <div>USB powered. Both disks are portable (one Toshiba
              Canvio, the other Seagate Expansion.<br>
               <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">
              - Intel, AMD? -- Intel 3Gbps chipset is better than AMD
              6Gbps.  </blockquote>
            <div><br>
            </div>
            <div>This is an AMD machine.<br>
               <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">You see the difference
              under heavy load, like 4 VMs in </blockquote>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">Hyper-V as I'm doing
              now.<br>
            </blockquote>
            <div><br>
            </div>
            <div>Bare metal in my case, no virtualization, and no other
              load on the machine.<br>
              <br>
            </div>
            <div>Seems there is a partial answer here:<br>
              <br>
              <a moz-do-not-send="true"
href="http://www.macworld.com/article/2039427/how-fast-is-usb-3-0-really-.html">http://www.macworld.com/article/2039427/how-fast-is-usb-3-0-really-.html</a><br>
              <br>
            </div>
            <div>Spinning disks can only do 114MB/s (7200 RPM that is).
              SSDs can do 200 MB/s.<br clear="all">
            </div>
          </div>
          <br>
        </div>
      </div>
    </blockquote>
    59MB/s sequential read speed is entirely reasonable for a 5400RPM
    drive - regardless of how it is connected.<br>
    <br>
    In order to get past the 200MB/s limit with a typical USB3
    connection, one has to be more creative eg.<br>
<a class="moz-txt-link-freetext" href="http://marc-abramowitz.com/archives/2007/02/17/getting-good-performance-out-of-usb-hard-drives-in-linux/">http://marc-abramowitz.com/archives/2007/02/17/getting-good-performance-out-of-usb-hard-drives-in-linux/</a><br>
    Tuning max_sectors should reduce CPU load (probably your bottleneck
    for fast USB transfer - watch "top") and increase sequential
    performance efficiency.<br>
    <br>
    <br>
    <br>
    <br>
    <br>
  </body>
</html>