ESCAPE 2000

ESCAPE 2000

eSCape2000 will explore the potential effects of the convergence of personal computing devices, the software infrastructure necessary to support them, and wireless connectivity and what it means to do "HPC Anywhere" in an event called eSCape 2000. The conference will provide a wireless infrastructure as part of SCinet 2000 to serve as a showcase for demonstrations of novel interfaces to HPC resources, data and applications. The tone of the event is that of a demonstration-not a competition-and the goal is to provide a forum for stimulating broad community discussions about how these new technologies will impact HPC. Participants are invited to demonstrate applications and technologies that they feel best illustrate how being able to reach "HPC Anywhere" will impact supercomputing and the people and industries that depend on supercomputing.

The event will be organized as a combination forum/permanent exhibit. The eSCape 2000 events will be held on the exhibit floor within a reserved section co-located with Research Gems. A presenter from each project will be available Wednesday and Thursday from 10:00 - 11:00 AM during exhibit hours for a presentation and demonstration of their research project. In addition, each team will be provided space within the eSCape 2000 venue to set up a permanent poster describing the project. The poster exhibit will remain up for the duration of the exhibit to provide those who cannot attend one of the presentations the opportunity to learn about the project.

Questions: escape2000@sc2000.org


  ESCAPE 2000 CO-CHAIRS
STEPHEN JONES, U.S. ARMY ENGINEER RESEARCH AND DEVELOPMENT CENTER
JOHN WEST, WORLDCOM


ESCAPE 2000 DEMONSTRATIONS

SDSC GridPort Toolkit Applications
Stephen Mock, SDSC/UCSD, Mary Thomas, Scientific Computing Department, SDSC

We will be demonstrating wireless access to applications that have been built using the SDSC GridPort Toolkit (https://hotpage.npaci.edu). These applications include the NPACI HotPage (https://hotpage.npaci.edu) and other computational science web portals. Access will allow users to view current HPC system information, including machine and job status.

The PUNCH Computing Portal: Integrating Grid Services and Web Technologies for High-Performance Computing Anytime, Anywhere
Jose Fortes, Rudolf Eigenmann, Mark Lundstrom, Purdue University, Valerie Taylor, Northwestern University, Miron Livny, University of Wisconsin-Madison, Sumalatha Adabala, Renato Figueiredo, Nirav Kapadia, Purdue University, Jose Miguel-Alonso, UPV, Spain

The demonstration will highlight computing portal and wide-area network-computing technologies that allow seamless management of high-performance applications, data and machines distributed across wide-area networks. These technologies allow users to access and run applications (even unmodified commercial applications) from anywhere via standard Web browsers. The presentation will outline the manner in which state-of-the-art technologies can be used to build end-to-end solutions for high-performance, ubiquitous supercomputing on Intranets and the Internet. The demonstration will be based on PUNCH, a network computing system that has been operational for five years. PUNCH currently provides computing services to about 800 users across ten countries; 50 engineering software packages developed by 13 universities and six vendors are available. PUNCH can be accessed at www.ece.purdue.edu/punch. The presentation will include an overview of the utility and benefits of such technologies in the context of several multi-university projects for collaborative engineering and distance education.

Speech-to-Text & Speech-to-Sign eSCapes Hearing Barriers Gottfried E. Zimmermann, Gregg Vanderheiden, University of Wisconsin-Madison Trace R&D Center, Dan Deignan, Personal Captioning

If you can't hear, it can be hard to participate in time-critical discussions. This can be because you are on a noisy factory floor, or because of functional impairment in hearing. Networked services, and particularly mobile unWired services could deliver more pervasive access to help in this area. Last year we demonstrated using Internet2 to make remote sign language interpreting feasible. This year we are experimenting with options for real-time transcription of speech to text and translation from speech to sign for collaborative participation of people with hearing impairments. In addition to exploring the capabilities of various strategies in and around the convention, we will demonstrate the techniques we are investigating to visitors at the eSCape2000 area in the exhibition hall.

Middleware and the eSCaped Web
Alfred S. Gilman, Gottfried E. Zimmermann, University of Wisconsin-Madison, Trace R&D Center

Network middleware to adapt web content for mobile devices is a very dynamic area. New players are emerging on all sides. There has been work in this area for the purpose of disability access for several years, and the W3C Note on guidelines for mobile-usable documents relies heavily on the Web Content Accessibility Guidelines 1.0 Recommendation developed for disability access purposes. We will be demonstrating how various middleware strategies perform against two usability criteria: fitness for use with small mobile devices, and fitness for use by people with disabilities.

The IU Enotebook
Dennis Gannon, Randall Bramley, Madhu Govindaraju, Benjamin Temko, Ken Chiu, Juan Villacis, Indiana University

The IU Enotebook allows access to and control of Grid resources from any browser-equipped device using only standard Web protocols. Features of the Enotebook include:

  1. Authentication via GSSAPI and Globus certificates of authority
  2. A Python scripting mechanism for locating, instantiating, connecting, and controlling Common Component Architecture HPC components across the grid
  3. A dual-level user model which includes script developers and end-users who select and run scripts for experiments
  4. Java servlet-based framework to allow users to disconnect and later reattach to running experiments from different IP addresses
  5. Tools for automatically handling files in a distributed computation, and providing a user with a full record of their experiments
  6. Component managers for rapid encapsulation and use of existing binaries which cannot be recompiled or directly linked with a component framework
  7. Support for multiple languages and communications protocols.