<node id="612011">
  <nid>612011</nid>
  <type>news</type>
  <uid>
    <user id="34540"><![CDATA[34540]]></user>
  </uid>
  <created>1537994031</created>
  <changed>1537994031</changed>
  <title><![CDATA[School of CSE Speeds up Graph Applications and Presents GraphChallenge Winning Paper at HPEC 2018]]></title>
  <body><![CDATA[<p>Georgia Tech&rsquo;s&nbsp;<a href="https://www.cse.gatech.edu/">School of Computational Science and Engineering</a>&nbsp;(CSE) presented six technical papers at the twenty-second annual&nbsp;<a href="http://www.ieee-hpec.org/">IEEE High Performance Extreme Computing</a>&nbsp;(HPEC 2018) conference this week.&nbsp;</p>

<p>Of the papers presented, three won or placed in a group of HPEC conference challenges, also referred to as&nbsp;the&nbsp;<a href="http://graphchallenge.mit.edu/champions">GraphChallenge</a>.</p>

<p>The GraphChallenge seeks to develop tests from diverse research communities and groundbreaking efforts such as the&nbsp;<a href="http://graph500.org/">Graph500</a>&nbsp;and&nbsp;<a href="http://graphanalysis.org/index.html">GraphAnalysis</a>, both of which CSE Chair&nbsp;<strong>David Bader&nbsp;</strong>is a committee member.&nbsp;</p>

<p><em><a href="http://tda.gatech.edu/papers/Yasar18-HPEC.pdf">Fast Triangle Counting Using Cilk</a>&nbsp;</em>by CSE Professor&nbsp;<strong>&Uuml;mit &Ccedil;ataly&uuml;rek&nbsp;</strong>and Ph.D. Student&nbsp;<strong>Abdurrahman Yasar</strong>, as well as researchers form Sandia National Laboratory, was one of four submissions to win the title of 2018 GraphChallenge champions.&nbsp;</p>

<p>&quot;Triangle counting is a representative graph analysis algorithm with several applications and one of three benchmarks used in the IEEE HPEC GraphChallenge,&rdquo; said Yasar. &ldquo;This paper improves upon the work that is implemented in&nbsp;<a href="https://github.com/kokkos/kokkos-kernels/wiki/Kokkoskernels">Kokkoskernels library</a>, one of the fastest implementations for the triangle counting problem in last year&rsquo;s GraphChallenge.&rdquo;</p>

<p>&ldquo;In addition to fast implementation, this work also provides a significantly improved analysis of the results, showing better agreement between theory and experiment,&rdquo; Yasar said.</p>

<p>Of the three winning submissions,&nbsp;<em>Fast and Adaptive List Intersections on the GPU</em>placed as a GraphChallenge Finalist Paper and&nbsp;<em><a href="https://www.researchgate.net/publication/327569911_Logarithmic_Radix_Binning_and_Vectorized_Triangle_Counting">Logarithmic Radix Binning and Vectorized Triangle Counting</a>&nbsp;</em>received an innovation award.&nbsp;</p>

<p>&quot;HPEC GraphChallenge provides a standard set of key graph problems that the larger community can work on. One of the difficulties is achieving good parallel performance across a wide variety of datasets,&rdquo; said CSE Ph.D. Student&nbsp;<strong>James Fox</strong>, lead investigator of the GraphChallenge finalist paper submission.&nbsp;</p>

<p>&ldquo;This paper contributes an input-adaptive approach to load-balancing and improving the performance of list intersections, important in triangle counting and other applications that involve finding common neighbors.&quot;&nbsp;</p>

<p>The conference takes place Sept. 25 through 27 in Waltham, Massachusetts.</p>

<p>&nbsp;</p>

<p><strong>Below are the titles of Georgia Tech&rsquo;s research being presented this week.</strong></p>

<ul>
	<li><em>Fast and Adaptive List Intersections on the GPU</em>&nbsp;- James Fox, Oded Green, Kasimir Gabert,&nbsp;Xiaojing An, David Bader&nbsp;</li>
	<li><em><a href="http://tda.gatech.edu/papers/Yasar18-HPEC.pdf">Fast Triangle Counting Using Cilk</a>&nbsp;-&nbsp;</em>Abdurrahman Yasar, Sivasankaran Rajamanickam, Michael Wolf, Jonathan Berry, &Uuml;mit &Ccedil;ataly&uuml;rek</li>
	<li>&nbsp;<a href="https://www.researchgate.net/publication/327569911_Logarithmic_Radix_Binning_and_Vectorized_Triangle_Counting"><em>Logarithmic Radix Binning and Vectorized Triangle Counting</em>&nbsp;</a>- Oded Green, James Fox, Alex Watkins, Alok Tripathy, Kasimir Gabert, Euna Kim, An Xiaojing, Kumar Aatish, David Bader&nbsp;</li>
	<li><em><a href="http://graphchallenge.mit.edu/champions">Performance Effects of Backing Data Stores in Community Detection Algorithms</a>&nbsp;</em>-&nbsp;Rohit Thankachan, Brian Swenson, James Fairbanks</li>
	<li><em><a href="https://www.researchgate.net/publication/327569969_Scaling_Betweenness_Centrality_in_Dynamic_Graphs">Scaling Betweenness Centrality in Dynamic Graphs</a>&nbsp;-</em>&nbsp;Alok Tripathy, Oded Green</li>
	<li><em><a href="http://on-demand.gputechconf.com/gtc/2018/presentation/s8177-hornet-efficient-data-structure-dynamic-sparse-graphs-matrices.pdf">Hornet: An Efficient Data Structure for Dynamic Sparse Graphs and Matrices on GPUs</a>&nbsp;-&nbsp;</em>Federico Busato, Oded Green, Nicola Bombieri, David Bader</li>
</ul>
]]></body>
  <field_subtitle>
    <item>
      <value><![CDATA[]]></value>
    </item>
  </field_subtitle>
  <field_dateline>
    <item>
      <value>2018-09-26T00:00:00-04:00</value>
      <timezone><![CDATA[America/New_York]]></timezone>
    </item>
  </field_dateline>
  <field_summary_sentence>
    <item>
      <value><![CDATA[CSE Researchers presented six technical papers with three winning GraphChallenge submissions to HPEC 2018.]]></value>
    </item>
  </field_summary_sentence>
  <field_summary>
    <item>
      <value><![CDATA[]]></value>
    </item>
  </field_summary>
  <field_media>
          <item>
        <nid>
          <node id="612009">
            <nid>612009</nid>
            <type>image</type>
            <title><![CDATA[HPEC Logo]]></title>
            <body><![CDATA[]]></body>
                          <field_image>
                <item>
                  <fid>232985</fid>
                  <filename><![CDATA[HPECLogo.png]]></filename>
                  <filepath><![CDATA[/sites/default/files/images/HPECLogo.png]]></filepath>
                  <file_full_path><![CDATA[http://tlwarc.hg.gatech.edu//sites/default/files/images/HPECLogo.png]]></file_full_path>
                  <filemime>image/png</filemime>
                  <image_740><![CDATA[]]></image_740>
                  <image_alt><![CDATA[]]></image_alt>
                </item>
              </field_image>
            
                      </node>
        </nid>
      </item>
      </field_media>
  <field_contact_email>
    <item>
      <email><![CDATA[kristen.perez@cc.gatehc.edu]]></email>
    </item>
  </field_contact_email>
  <field_location>
    <item>
      <value><![CDATA[]]></value>
    </item>
  </field_location>
  <field_contact>
    <item>
      <value><![CDATA[<p><strong>Kristen Perez</strong></p>

<p>Communications Officer I</p>

<p>College of Computing - School of Computational Science and Engineering</p>
]]></value>
    </item>
  </field_contact>
  <field_sidebar>
    <item>
      <value><![CDATA[]]></value>
    </item>
  </field_sidebar>
  <field_boilerplate>
    <item>
      <nid><![CDATA[]]></nid>
    </item>
  </field_boilerplate>
  <!--  TO DO: correct to not conflate categories and news room topics  -->
  <!--  Disquisition: it's funny how I write these TODOs and then never
         revisit them. It's as though the act of writing the thing down frees me
         from the responsibility to actually solve the problem. But what can I
         say? There are more problems than there's time to solve.  -->
  <links_related> </links_related>
  <files> </files>
  <og_groups>
          <item>47223</item>
          <item>50877</item>
      </og_groups>
  <og_groups_both>
      </og_groups_both>
  <field_categories>
      </field_categories>
  <core_research_areas>
      </core_research_areas>
  <field_news_room_topics>
      </field_news_room_topics>
  <links_related>
      </links_related>
  <files>
      </files>
  <og_groups>
          <item>47223</item>
          <item>50877</item>
      </og_groups>
  <og_groups_both>
          <item><![CDATA[College of Computing]]></item>
          <item><![CDATA[School of Computational Science and Engineering]]></item>
      </og_groups_both>
  <field_keywords>
          <item>
        <tid>4305</tid>
        <value><![CDATA[cse]]></value>
      </item>
          <item>
        <tid>13255</tid>
        <value><![CDATA[david bader]]></value>
      </item>
          <item>
        <tid>3427</tid>
        <value><![CDATA[High performance computing]]></value>
      </item>
          <item>
        <tid>170627</tid>
        <value><![CDATA[Umit Catalyurek]]></value>
      </item>
      </field_keywords>
  <field_userdata>
      <![CDATA[]]>
  </field_userdata>
</node>
