<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Timestamp counters on multiple processors in Software Archive</title>
    <link>https://community.intel.com/t5/Software-Archive/Timestamp-counters-on-multiple-processors/m-p/994746#M28109</link>
    <description>&lt;DIV&gt;Can timestamp counters on multiple processors be relied on to be synchronous at any time? I. e. whether they start simultaneously on systeminitialization and whether their spin can be affected by processor state  SpeedStep etc. Question concerns both different logical processors in the same package as well as different physical processors in a system. Thanks.&lt;/DIV&gt;</description>
    <pubDate>Tue, 01 Mar 2005 03:31:27 GMT</pubDate>
    <dc:creator>dm71</dc:creator>
    <dc:date>2005-03-01T03:31:27Z</dc:date>
    <item>
      <title>Timestamp counters on multiple processors</title>
      <link>https://community.intel.com/t5/Software-Archive/Timestamp-counters-on-multiple-processors/m-p/994746#M28109</link>
      <description>&lt;DIV&gt;Can timestamp counters on multiple processors be relied on to be synchronous at any time? I. e. whether they start simultaneously on systeminitialization and whether their spin can be affected by processor state  SpeedStep etc. Question concerns both different logical processors in the same package as well as different physical processors in a system. Thanks.&lt;/DIV&gt;</description>
      <pubDate>Tue, 01 Mar 2005 03:31:27 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/Timestamp-counters-on-multiple-processors/m-p/994746#M28109</guid>
      <dc:creator>dm71</dc:creator>
      <dc:date>2005-03-01T03:31:27Z</dc:date>
    </item>
    <item>
      <title>Re: Timestamp counters on multiple processors</title>
      <link>https://community.intel.com/t5/Software-Archive/Timestamp-counters-on-multiple-processors/m-p/994747#M28110</link>
      <description>I will oversimplify, to avoid overstepping my limited expertise:&lt;BR /&gt;&lt;BR /&gt;Within a single CPU package, the internal logic has the responsibility of keeping time stamp counters synchronized.  This restricts freedom of individual components to do power saving state changes.&lt;BR /&gt;&lt;BR /&gt;With multiple CPU packages on the same motherboard, the chip set logic has a similar responsibility, but there may be future platforms which don't take care of this.&lt;BR /&gt;&lt;BR /&gt;New power management features require a means of making time stamp counting continuous across routine state changes, but I don't think there is any simplified generalization to tell which state changes will break it.&lt;BR /&gt;&lt;BR /&gt;If your requirements are more exacting, the details probably exceed what can be discussed here, and may involve going through your account representative.</description>
      <pubDate>Tue, 01 Mar 2005 05:21:39 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/Timestamp-counters-on-multiple-processors/m-p/994747#M28110</guid>
      <dc:creator>TimP</dc:creator>
      <dc:date>2005-03-01T05:21:39Z</dc:date>
    </item>
  </channel>
</rss>

