<?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 Hi Zvi, in OpenCL* for CPU</title>
    <link>https://community.intel.com/t5/OpenCL-for-CPU/Consumer-Kernel-with-pipe/m-p/1100263#M5109</link>
    <description>&lt;P&gt;Hi Zvi,&lt;/P&gt;

&lt;P&gt;We don't have concurrent kernels in the production driver, so producer kernel has to finish first and then consumer kernel will start. As you can imagine, this is not terribly useful, that's why I don't recommend currently to use pipes. But, to answer your question, in our current implementation of the driver it cannot happen for the above reason.&lt;/P&gt;</description>
    <pubDate>Fri, 22 Jan 2016 21:50:57 GMT</pubDate>
    <dc:creator>Robert_I_Intel</dc:creator>
    <dc:date>2016-01-22T21:50:57Z</dc:date>
    <item>
      <title>Consumer Kernel with pipe</title>
      <link>https://community.intel.com/t5/OpenCL-for-CPU/Consumer-Kernel-with-pipe/m-p/1100262#M5108</link>
      <description>&lt;P&gt;Hello,&lt;/P&gt;

&lt;P&gt;The following code is in a "consumer kernel":&lt;/P&gt;

&lt;PRE class="brush:cpp;"&gt;//reserve pipe for reading
reserve_id_t rid = work_group_reserve_read_pipe(rng_pipe, szgr);
if(is_valid_reserve_id(rid)) {
&amp;nbsp; //read random number from the pipe.
&amp;nbsp; read_pipe(rng_pipe,rid,lid, &amp;amp;rn);
&amp;nbsp; work_group_commit_read_pipe(rng_pipe, rid);
&amp;nbsp; }&lt;/PRE&gt;

&lt;P&gt;Is it possible that read_pipe will be blocked till it gets a vector of data written from a producer kernel ?&lt;/P&gt;

&lt;P&gt;Regards,&lt;/P&gt;

&lt;P&gt;Z.V&lt;/P&gt;</description>
      <pubDate>Fri, 22 Jan 2016 21:43:15 GMT</pubDate>
      <guid>https://community.intel.com/t5/OpenCL-for-CPU/Consumer-Kernel-with-pipe/m-p/1100262#M5108</guid>
      <dc:creator>ZVere</dc:creator>
      <dc:date>2016-01-22T21:43:15Z</dc:date>
    </item>
    <item>
      <title>Hi Zvi,</title>
      <link>https://community.intel.com/t5/OpenCL-for-CPU/Consumer-Kernel-with-pipe/m-p/1100263#M5109</link>
      <description>&lt;P&gt;Hi Zvi,&lt;/P&gt;

&lt;P&gt;We don't have concurrent kernels in the production driver, so producer kernel has to finish first and then consumer kernel will start. As you can imagine, this is not terribly useful, that's why I don't recommend currently to use pipes. But, to answer your question, in our current implementation of the driver it cannot happen for the above reason.&lt;/P&gt;</description>
      <pubDate>Fri, 22 Jan 2016 21:50:57 GMT</pubDate>
      <guid>https://community.intel.com/t5/OpenCL-for-CPU/Consumer-Kernel-with-pipe/m-p/1100263#M5109</guid>
      <dc:creator>Robert_I_Intel</dc:creator>
      <dc:date>2016-01-22T21:50:57Z</dc:date>
    </item>
  </channel>
</rss>

