<html>
Hello, <br><br>
Attached in MS Word and copied below is an NCSA Seminar we will be
broadcasting via Access Grid on Thursday, October 3, 2002 at 4-5pm
central using the NCSA system. I would like to have this listed on
your calendar. If you have any questions, please don't hesitate to
contact me at 217-244-0078 or via email.<br><br>
thanks for your help,<br>
Beth McKown<br>
*******************<br>
<div align="center"><font size=5>Green Destiny: A 240-Node Compute
Cluster in One Cubic Meter<br><br>
Wu-chun Feng<br>
Los Alamos National Laboratory<br><br>
Thursday, October 3, 2002<br>
5602 Beckman Institute<br>
Also via Access Grid
<a href="http://agschedule.ncsa.uiuc.edu/" eudora="autourl"><b>http://agschedule.ncsa.uiuc.edu</a>
</b>4:00pm<br><br>
</font></div>
<b><i>Abstract<br>
</i></b>In this talk, Dr. Feng will present a new twist to the Beowulf cluster the Bladed Beowulf. In contrast to traditional Beowulfs, Dr. Feng's Bladed Beowulf, dubbed Green Destiny, uses Transmeta processors in order to keep thermal power dissipation low and reliability and density high while still achieving comparable performance to Intel- and AMD-based clusters when their processors run at the same clock rate.<br><br>
Although performance and price/performance have been (and will continue to be) important metrics in high-performance computing, Dr. Feng believes that the issues of efficiency, reliability, and availability will be the key issues of this decade. Bigger and faster machines will not be good enough anymore. Consequently, he proposes a new performance metric, Total Price/Performance Ratio (ToPPeR), to evaluate high-performance computing systems. However, due to the hidden and institution-specific costs in the ToPPeR metric, two alternative (but more concrete) performance metrics that are related to efficiency, reliability, and availability are also proposed: performance/power ratio ("power efficiency") and performance/space ratio ("space efficiency"). Dr. Feng will then present empirical data for the aforementioned metrics over a variety of computing platforms using a cosmology code and a bioinformatics code. <br><br>
For more information about this project, please visit the ^ÓSupercomputing in Small Spaces^Ô web site at <a href="http://sss.lanl.gov/" eudora="autourl"><font color="#0000FF"><u>http://sss.lanl.gov</a></u></font>. For media coverage on the project, see CNN^Òs coverage at <a href="http://www.cnn.com/2002/TECH/industry/05/21/supercomputing.future.idg/index.html" eudora="autourl"><font color="#0000FF"><u>http://www.cnn.com/2002/TECH/industry/05/21/supercomputing.future.idg/index.html<br>
</a></u></font>or the New York Times at <a href="http://www.nytimes.com/2002/06/25/science/physical/25COMP.html" eudora="autourl"><font color="#0000FF"><u>http://www.nytimes.com/2002/06/25/science/physical/25COMP.html</a></u></font>.<br><br>
<b><i>Biography<br>
</i></b>Dr. Feng received a B.S. degree in computer engineering and a B.S. (Honors) degree in music from Penn State University in 1988; an M.S. degree in computer engineering from Penn State University in 1990; and a Ph.D. degree in computer science from the University of Illinois at Urbana-Champaign in 1996. <br><br>
He is currently a technical staff member and team leader of RADIANT (Research & Development in Advanced Network Technology at http://www.lanl.gov/radiant) at Los Alamos National Laboratory and an adjunct assistant professor at the Ohio State University. He is a fellow of the Los Alamos Computer Science Institute and the founder and director of the Advanced Summer Curriculum for Emerging Network Technologies (ASCENT). Before joining LANL in 1998, Dr. Feng had previous professional stints at Purdue University, University of Illinois, NASA Ames Research Center, and IBM T.J. Watson Research Center. <br>
</div>
</html>