US National Labs Salishan Fellowship

Salishan | Algorithms, Architecture, Language

Los Alamos | Lawrence Livermore | Lawrence Berkeley | Sandia

Christopher Altman  US graduate student in applied physics, Kavli Institute of Nanoscience, Delft University of Technology. Research foci include high-performance computing (HTMT), solid-state superconducting nanoelectronics, macroscopic quantum coherence and computation, quantum infomation processing in Josephson junction nanocircuits.

I was honored to participate in the US National Labs Conference on High Speed Computing from April 23-27, 2001 at Westin Salishan in Gleneden Beach, Oregon. Salishan is a half-mile's walk from the beach, a picturesque, mist-covered mountain resort that has been the setting for the conference since its inception.

The conference, founded in 1980, was founded as a means of getting experts in computer architecture, languages, and algorithms together to improve communications, develop collaborations, solve problems of mutual interest, and provide effective leadership in the field of high speed computing. Attendance is by invitation only, and limited to about 170 of the best and brightest in the world.

The conference is sponsored by Lawrence Livermore, Los Alamos, and Sandia National Laboratories, as well as being co-sponsored by a number private companies—this year volunteering sponsors included Compaq, Cray Inc., Fujitsu, IBM, Intel, SAIC, SGI, StorageTek, and Sun Microsystems.

A highlight of the conference was the informal discussions held each evening in Salishan's Sunset Suite, a forum to exchange ideas, solve problems, and develop friendships. This year's talks profiled recent developments in nanotechnology, supercomputing, microelectromechanical systems, large-scale networks, memory architectures, data management, artificial intelligence, molecular electronics, and a number of other technologies that will significantly impact the future of information science and technology.

The meeting was a stimulating and challenging week of close interaction with many of the most creative minds in the field. I'd like to extend my gratitude to the many inspiring scientists with whom I had the opportunity to meet at Salishan—and to those who helped to make my attendance possible, including Doc Bedard, Horst Simon, David Kahaner, Brett Berlin, Will Stackhouse, Jim McGraw, Kathy Turnbeaugh, Dennis Bohnenkamp, and Lala Stone.

It was an honor to attend under support of a Salishan fellowship, to meet and discuss large-scale networks with H. Shrikumar and paintable computers with Bill Butera. I look forward to meeting both again on my next trip to the Media Lab.

Special thanks go to Horst Simon for our continuing discussions on high-performance computing and the HTMT architecture, and to Doc Bedard for his guidance, advice, and for answering my questions while exploring ideas on imaginative walks through the forested grounds of Salishan. My interest in Josephson Junction RSFQ superconducting nanoelectronics has been in no small part due to Bedard's Random Access talk, and the influence of our continued discussions throughout the week.

Notwithstanding revolutionary hardware breakthroughs, the next generation of high-performance computing systems will continue to be reliant upon low-temperature superconducting nanoelectronics. Moore's Law ensures their dimensions will shrink rapidly. As we enter the era of quantum information processing, this is certain to be a productive and exciting area of research.


Application Requirements and Current System Architectures

An Overview of Nuclear Stockpile Stewardship
James Mercer-Smith, Los Alamos National Laboratory

Requirements for Large-Scale Massively Parallel Computing
Robert Weaver, Los Alamos National Laboratory

Sandia C-Plant Clusters
Art Hale, Sandia National Laboratories

LLNL ASCI Platforms
Mark Seager, Lawrence Livermore National Laboratory

Future System Architectures

HEC Architectures in the 21st Century: Drivers and Imperatives
Thomas Sterling, Caltech

High Performance and High Density Archives
Jim Hughes, StorageTek

Future Communications and Networking
Marc Beackon, Lucent Technologies

From Problem Definition to Problem Setup

An Introduction to the Challenges of Problem Setup
Robert Leland, Sandia National Labs

Responses to Analysis / CAD Integration Perplexities
Ted Blacker, Fluent, Inc.

Computational Problem Setup: An Industrial Perspective
Todd Michal, Boeing

Unstructured Meshing
Glen Hanson, Los Alamos National Laboratory

A Heirarchical Data Management System for Parallel Partitioning of Adaptive Communication
Joe Flaherty, Renssalaer Polytechnic University

CAD to Results: The Snowball Effect
David White, Cargegie-Mellon University

From Problem Setup to Result Data

Performance Metrics: Out of the Dark Ages
David Bailey, Lawrence Berkeley National Laboratory

State of the Art in Programming Tools
John Levesque, Times N Systems

Addressing the Memory Bottleneck
Sally McKee, University of Utah

Random Access Talks

NASA's digital library initiative
Eugene Miya, NASA Ames Research Center

Evolutionary Hardware
de Garis, Starlab NV/SA

OSCAR and the Open Cluster Group

ASCI Setup
Sandia National Laboratory

Self-adapting software
Jim Hughes, INFOSEC

Josephson Junction RSFQ Superconducting nanoelectronics
Fernand Bedard, National Security Agency

ATIP Activities in East Asia
David Kahaner, Asian Technology Information Program

Alan Huang, Stanford University

Norm Whittaker

From Result Data to Insight

Is Visualization a Solved Problem?
Sam Uselton, Lawrence Livermore National Laboratory

Large Scale Scientific Data Management and Analysis
Alok Choudhary, Northwestern University

Can Data Mining Ever be a Gigabit Application? Lessons from DataSpace
Robert Grossmsan, University of Illinois, Chicago

Schooling in the Digital Age
Sara Armstrong, PhD, George Lucas Educational Foundation

MEMS: Micro-Electrical Mechanical Systems

A Smaller Hammer
William S. Trimmer, Standard MEMS Inc.

MEMS Modeling: Pushing the Limits of Miniaturization
Robert Rudd, Lawrence Livermore National Laboratory

Artificial Brains and Self-Configuring Electronics

Artificial Brains: Today and Tomorrow
de Garis, Starlab NV/SA

Gate Array, Configure Thyself
Nick Macias, Cell Matrix Corp

An Approach to Designing Extremely Large, Extremely Parallel Systems
Lisa Durbeck, Cell Matrix Corp

Molecular Computing and Myriad Nets

Defect Tolerant Molecular Electronics Algorithms, Architectures, and Atoms
Philip Kuekes, Hewlett-Packard Laboratories

Myriad Nets: De-Layering to Scale Networks up to the Billions
H. Shrikumar, MIT Media Laboratory

Future Directions

Programming a Paintable Computer
Bill Butera, MIT Media Laboratory

The Future of High Performance Computing: Dynamic Translation and High Density Computing
Dave Taylor, Transmeta Corporation

No comments:

Post a Comment