ColloquiaTalk by Ken Kennedy, Rice University - April 2nd

[Margery Ishmael]

Tuesday, April 2, 2002
3:00 pm
Ryerson 251

Ken Kennedy, Director for High Performance Software, Rice University

"Telescoping Languages: A Framework for Generating High Performance 
Problem-Solving Systems"

Abstract: Increases in the complexity of both computer architectures and 
application structure have led to corresponding increases in the difficulty 
of application development, making it the nearly exclusive domain of the 
professional programmer.  When combined with the current shortage of 
programmers at the highest skill levels, this has created a software gap 
between the demand for new software and the ability of our work force to 
deliver it.

One way to bridge this gap is to make it possible for end users to develop 
applications for themselves.  Indeed, many users today are producing highly 
functional applications using scripting languages and high-level 
problem-solving systems such at Matlab and Visual Basic.  Unfortunately, 
these applications are not typically considered in the statistics that 
measure productivity because they fall short in some measure of 
performance.  Any such application that is deemed useful is usually 
rewritten in a more traditional programming language - C, C++, or Fortran - 
before being used for "production" work.  If we could skip this rewriting 
step, we would take a significant step toward overcoming the software gap, 
while making it possible for end users to develop truly powerful 
high-performance applications.

At Rice we are developing a framework, called telescoping languages, for 
generating optimized high-level problem-solving languages from annotated 
domain libraries.  The strategy involves an extensive, compute-intensive 
preliminary analysis of the library, performed  at language-generation 
time.  The output of this process, which could take many hours to complete, 
will be an efficient compiler for an extended scripted language in which 
calls to the underlying domain library are recognized and optimized as 
primitive operations.  The talk will describe this strategy and its 
applications in detail and report on some preliminary experiments 
demonstrating its effectiveness.

Bio: Ken Kennedy is an Ann and John Doerr Professor in Computational 
Engineering at Rice University's Computer Science Department.  In addition, 
he is the director of HiPerSoft, a Center for High Performance Software 
Research headquartered at Rice University, and is also a Co-Chair of PITAC, 
the President's Information Technology Advisory Committee. 
http://www.cs.rice.edu/~ken/

Hosts: Ridgway Scott & Rick Stevens

*The talk will be followed by refreshments in Ryerson 255*
Persons with disabilities who may need assistance should call 773.834.8977


=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Margery Ishmael
Secretary to the Chairman, Department of Computer Science
The University of Chicago
1100 E. 58th Street, Chicago, IL. 60637-1581
tel. 773.834.8977  fax. 773.702.8487
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=