Date: Wed, 20 Nov 1996 19:33:58 GMT
Server: Apache/1.1.1
Content-type: text/html
Content-length: 4187
Last-modified: Tue, 20 Aug 1996 23:17:29 GMT

<title>   Numerical Computation Research Group</title><h3>    Computer Science Department<br>    University of Colorado at Boulder<br>    Academic Year 1996/97</h3><hr><br><h1>   Numerical Computation Research Group</h1><h2>   Faculty</h2>   <!WA0><a href="http://www.cs.colorado.edu/~lizb/Home.html">      Liz Bradley   </a>,   <!WA1><a href="http://www.cs.colorado.edu/~richard/Home.html">      Richard Byrd   </a>,   <!WA2><a href="http://www.cs.colorado.edu/~cai/Home.html">      Xiao-Chuan Cai   </a>,   <!WA3><a href="http://www.cs.colorado.edu/~jessup/Home.html">      Liz Jessup    </a>,   <!WA4><a href="http://www.cs.colorado.edu/~mcbryan/Home.html">      Oliver McBryan   </a>,   <!WA5><a href="http://www.cs.colorado.edu/~bobby/Home.html">      Bobby Schnabel   </a>.<h2>   Research Interests </h2><p>The numerical computation research group is concerned withdeveloping, analyzing, implementing, and testing methods forsolving mathematical problems on computers.  It is alsoconcerned with applications of numerical methods to scientificand engineering applications, and with languages and tools for numerical computation.  There is considerable emphasison the development of methods for parallel computers.<p>Specific areas of emphasis within the group include optimizationand nonlinear equations (Byrd, Schnabel), partial differentialequations (Cai, McBryan), eigenvalue problems (Jessup), andlanguages and tools for parallel computation (Jessup, Schnabel).Currently there is also considerable emphasis on applicationsfrom molecular chemistry (Byrd, Schnabel), astrophysics (McBryan),space structures (McBryan), and combustion and chaotic systems (Bradley).<p>Courses and a textbook in<!WA6><a href = "http://www.cs.colorado.edu/ftp/pub/HPSC/README.html>   High Performance Scientific Computing</a>have been developed for both undergraduates and graduate students.A specially equipped laboratory, the HPSC lab, provides computing supportfor these courses.<h2>   Major Funded Research Projects</h2>The main funded research projects in this area include:<ul>   <li> High performance computing issues for turbulence and aeroelastic           simulation -- funded by ARPA, NSF, and NASA, including an            NSF Grand Challenge grant and a NASA Grand Challenge grant   <li> Large-scale optimization methods, with emphasis on applications              from molecular chemistry -- funded by AFOSR, ARO, NSF   <li> Numerical methods for structured sparse matrix problems --           funded by DOE    <li> Domain decomposition methods for partial differential equations  --           funded by NSF</ul><p>In addition, each faculty member in this group who was eligibleto compete for an NSF Young Investigator award -- Bradley, Cai,and Jessup -- has received this award. <h2>Professional Leadership Activities in Numerical Computation</h2>Selected activities in the numerical computation community are listedbelow.<ul>   <li> Byrd: Editorial board, Applied Mathematics Letters   <li> McBryan: Editorial board, IMPACT of Computing in Scientific           Applications, Supercomputing Review, Concurrency: Practice and            Experience; member, Army High Performance Computing Research Center           Board, NIST Computer Board, Los Alamos National Laboratory Computing           Review Board   <li> Schnabel: Past chair, ACM Special Interest Group on Numerical           Mathematics; Associate editor, Mathematical Programming,	   SIAM Review, SIAM Journal on Optimization;	   member, SIAM Council, RIACS Science Council; co-founder,	   SIAM Activity Group on Optimization</ul>   <h2>Facilities</h2>In addition to utilizing the department's many workstations,the numerical computation group makes extensive use of theparallel computing facilities available in the department.Currently these include:<ul>   <li> Intel iPSC-2 Distributed Memory Multiprocessor (32 processors)   <li> KSR1 Virtual Shared Memory Multiprocessor (64 processors)   <li> Intel Paragon Distributed Memory Multiprocessor            (208 nodes, 6GB memory), housed at NOAA Laboratory, Boulder </ul>