Francine Berman - US grants

Proper mapping of processors is a critical component of programming on message-passing machines. The proposed research focuses on the design and development of a prototype environment-sensitive mapping tool which allows the programmer to generate, assess and optimize mappings. The prototype mapping tool will be integrated with modified tools for parallel debugging, performance monitoring and program visualization. The resulting new more complete environment will be targeted to MIMD, message passing multiprocessors.//

9301788 Berman We propose to design and develop a tool for visualizing, displaying and optimizing the data and function partitioning of large programs coded in parallel languages with explicit decomposition constructs. This tool will assist in ensuring that the decomposition of processes and data will promote the most efficient use of parallelism for a given program and target multicomputer. Coordinated with this tool will be a parallel debugger (Panorama), which is built on the resident debugger of a commercial multicomputer, and which will provide a number of program views that will illuminate data movement and message traffic from the program execution.

9318817 Berman University of California, San Diego has just been given the opportunity to acquire from DoD (NRaD) an 8K CM-2. This machine is a SIMD machine and provides a distinctly different programming model than that of machines at UCSD and the San Diego Supercomputer Center (SDSC). Acquisition of the CM-2 would provide a major resource for our teaching and research. The CM-2 would be enormously useful to the Parallel Computation Lab, CSE Department and UCSD in three areas: teaching, research and heterogeneous computing.

9308900 Berman In this proposal, we outline a model and a mechanism for scheduling large-scale scientific applications on a non-dedicated heterogeneous network of computers. To obtain efficient schedules, our system will consider both domain-and application- specific information as well as system-wide characteristics. We seek a scalable and portable scheduling system with applicability to a wide variety of computational environments. We plan to use two important scientific applications with diverse computational requirements (global climate modeling and PACVHM) to test our scheduler. The initial implementation will use PVM as the underlying control mechanism. Our intent, is that the scheduler will be able to use any distributed task control as it foundation. ***

9423802 Berman Applicants to the NSF 1994 Minority Graduate Fellowship competition who were awarded "Honorable Mention" status and who enrolled in a computer science or computer engineering graduate program at a U.S. university were eligible to apply to the CISE Directorate for this special award. The purpose of the award is to assist the student in both research and educational activities related to his/her graduate education. The award is made on behalf of the student to the institution with the student's advisor designated as principal investigator. ***

9416785 Irwin This award to the Committee on the Status of Women (CRAW), provides for travel support for committee meetings, for continuing projects, and for new projects. CRAW, a committee of the Computing Research Association (CRA), was established in 1990. The goal of the Committee is to take positive action to increase the number of women participating in computer science and engineering research and education at all levels. Dr. Mary Jane Irwin, a Professor of Computer Science and Engineering at Pennsylvania State University, and Dr. Fran Berman, an Associate Professor of Computer Science at the University of California, San Diego are co-chairs of CRAW. CRAW has been very active in recommending and organizing projects and in many cases obtaining separate funding for them. Members of CRAW have produced several articles on women's issues in CS for the Computing Research News; have started the Systers-Academia, a moderated electronic mailing list for female faculty and Ph.D. students in CS&E; have established the Ph.D. database of women with Ph.D's in CS&E; and have presented career management workshops at major conferences and events (with additional funding). The CRA is an association of U.S. and Canadian academic departments of Computer Science and Computer Engineering, and Industrial Laboratories engaging in basic computing research. The mission of CRA is to represent and inform the computing research community and to support and promote its interests. ***

Fast networks have made it possible to aggregate distributed CPU, memory, storage, and data to provide the potential for application performance superior to that attainable on any single system. However, achieving performance on these metacomputing systems has proven difficult. Currently, application developers use customized application schedules to achieve performance on a metacomputer. Such schedules are based on a prediction of how the application will execute, given its resource requirements, the performance capabilities of the system resources, and the time-varying contention effects caused by competing applications executing on the metacomputer. This project's goal is to use the application-centric scheduling paradigm that is emerging from the practices of metacomputing applications developers as the basis for scheduling software which promotes application performance on metacomputing systems. To do this, the PIs will design and implement application-centric scheduling agents - called AppLeS (Application Level Scheduler) agents - for individual metacomputing applications. AppLeS agents will schedule the components of an individual parallel application on a non-dedicated system according to the dynamically changing performance each resource can deliver.

EIA-9820471
Computing Research Association
Berman, Francine
Aspray, William

Special Project: Travel Support for the CRA-W FCRC ``Super-Mentoring" Workshop: Atlanta GA
This grant aims at sponsoring ``super-mentoring'' workshop for FCRC
'99 which expands the previous (primarily junior and academic)
mentoring workshop target audience to include women involved in
research careers at national laboratories and within the industrial
sector, as well as senior (post-tenure) women who may be beginning to
deal with impediments to further promotion. The requested support is
for travel grants for participants and speakers, and support for the
primary workshop social event. Past experience has shown that the
travel grants are critical for many women who could not have attended
otherwise. Based on the comments from women who have attended similar
workshops organized by the Computing Research Association in the
past, it is expected that this workshop will serve as a catalyst for
positive and important changes in participant's lives.

Ensembles of distributed communication, computation, and storage resources, also known as "Computational Grids", are emerging as a critical platform for high-performance computing. Grids are used effectively to support runs of distributed applications at a large enough scale to provide new disciplinary results to their developers. Researchers in almost every field of science and engineering are particularly interested in a class of applications particularly well suited to the Grid, scientific simulations where many parameterized instances of a give computation are performed. The development of accessible, efficient, fault-tolerant Grid-enabled versions of simulation software will enable disciplinary scientists to investigate wide-ranging scenarios and to obtain new results orders of magnitude faster than is currently possible.

Many scientists would like to view large-scale simulations as software instruments that support some level of user interaction. This would be effective only if simulations can be deployed easily and controlled dynamically, i.e. if the computation can be steered. A traditional scenario is for the user to steer the simulation based on partial results that evolve continuously during execution. The partial results provide an increasingly refined indicator of the final results of the simulation and can be used to identify mid-execution which parameter sets are most promising. Given the potential of wide-area, federated Grid environments to deliver the aggregate computational power, data storage and dissemination facilities for large-scale simulations, and the need for scientists to steer such computations, it is increasingly important to develop performance-efficient and steerable software instruments that target the Grid. This project will address the significant computer science problems that arise from the need to support steerable scientific simulations in large-scale Grid environments.

The project will design, develop, and prototype a virtual software instrument as a vehicle for designing and prototyping scalable, steerable scientific simulations for the Grid. It will use a Monte Carlo simulation program, MCell, as a prototype application for development and testing of the virtual instrument. The virtual instrument itself will consist of a set of software modules, libraries, interfaces, and steering-sensitive scheduling algorithms. The project will have impact on both the computer science and disciplinary science communities. It will foster new research in computer science through the development of event models, performance models, data management strategies, and adaptive scheduling and steering algorithms. It will also enable domain scientists to obtain new results in neuroscience.

The University of California, San Diego, in collaboration with University of Illinois at Urbana-Champaign, is funded by a Cooperative Agreement from the Division of Advanced Computational Infrastructure in the amount of $26,450,500 to construct a Distributed Terascale Facility (DTF) that will span four institutions: 1) San Diego Supercomputer Center, 2)National Center for Supercomputing Applications , 3) California Institute of Technology and 4) Argonne National Laboratory . In partnership with IBM, Intel, Myrinet, Qwest, Oracle and SUN, they will construct a DTF based on multiple terascale Linux clusters, as well as large-scale storage archives and data management software. A 40 gigabit/second optical mesh will interconnect the DTF's components.

The DTF hardware will be integrated, using Globus middleware and other Computational Grid technologies develop by the Partnerships for Advanced Computational Infrastructure (PACI), to create a system with an aggregate of 11.6 TF of computing capability and 525 TB of disk storage to support terascale computing (6.1 TF and 153 TB at NCSA), distributed data management (4.1 TF and 248 TB at SDSC), remote rendering and visualization (1 TF and 31 TB at Argonne), and data-intensive scientific application analysis (0.4 TF and 93 TB at Caltech). A unified Operations Center will coordinate management, user support, and access.

The DTF TeraGrid will have broad impact on the computational science community and will ultimately benefit society as a whole. First, the TeraGrid will advance discovery and promote understanding by making available to academic researchers next generation information technologies that are an order of magnitude more capable than now generally available. Second, the TeraGrid will empower broad scientific and engineering communities through its focus on commodity clusters, common usage models and community codes. Third, experience gained through DTF operation, insights and software from the TeraGrid will be packaged in a series of software distributions for community use. This software packaging will extend TeraGrid technologies to the broader national community, including Minority Serving Institutions (MSI) and EPSCoR Institutions.

This collaborative project aims to establish a national computational resource to move the research community much closer to the realization of the goal of the Tree of Life initiative, namely, to reconstruct the evolutionary history of all organisms. This goal is the computational Grand Challenge of evolutionary biology. Current methods are limited to problems several orders of magnitude smaller, and they fail to provide sufficient accuracy at the high end of their range.

The planned resource will be designed as an incubator to promote the development of new ideas for this enormously challenging computational task; it will create a forum for experimentalists, computational biologists, and computer scientists to share data, compare methods, and analyze results, thereby speeding up tool development while also sustaining current biological research projects.

The resource will be composed of a large computational platform, a collection of interoperable high-performance software for phylogenetic analysis, and a large database of datasets, both real and simulated, and their analyses; it will be accessible through any Web browser by developers, researchers, and educators. The software, freely available in source form, will be usable on scales varying from laptops to high-performance, Grid-enabled, compute engines such as this project's platform, and will be packaged to be compatible with current popular tools. In order to build this resource, this collaborative project will support research programs in phyloinformatics (databases to store multilevel data with detailed annotations and to support complex, tree-oriented queries), in optimization algorithms, Bayesian inference, and symbolic manipulation for phylogeny reconstruction, and in simulation of branching evolution at the genomic level, all within the context of a virtual collaborative center.

Biology, and phylogeny in particular, have been almost completely redefined by modern information technology, both in terms of data acquisition and in terms of analysis. Phylogeneticists have formulated specific models and questions that can now be addressed using recent advances in database technology and optimization algorithms. The time is thus exactly right for a close collaboration of biologists and computer scientists to address the IT issues in phylogenetics, many of which call for novel approaches, due to a combination of combinatorial difficulty and overall scale. The project research team includes computer scientists working in databases, algorithm design, algorithm engineering, and high-performance computing, evolutionary biologists and systematists, bioinformaticians, and biostatisticians, with a history of successful collaboration and a record of fundamental contributions, to provide the required breadth and depth.

This project will bring together researchers from many areas and foster new types of collaborations and new styles of research in computational biology; moreover, the interaction of algorithms, databases, modeling, and biology will give new impetus and new directions in each area. It will help create the computational infrastructure that the research community will use over the next decades, as more whole genomes are sequenced and enough data are collected to attempt the inference of the Tree of Life. The project will help evolutionary biologists understand the mechanisms of evolution, the relationships among evolution, structure, and function of biomolecules, and a host of other research problems in biology, eventually leading to major progress in ecology, pharmaceutics, forensics, and security.

The project will publicize evolution, genomics, and bioinformatics through informal education programs at museum partners of the collaborating institutions. It also will motivate high-school students and college undergraduates to pursue careers in bioinformatics. The project provides an extraordinary opportunity to train students, both undergraduate and graduate, as well as postdoctoral researchers, in one of the most exciting interdisciplinary areas in science. The collaborating institutions serve a large number of underrepresented groups and are committed to increasing their participation in research.

This award funds a twelve-month planning effort to initiate technology training and support nationally to kindergarten through high school (K-12) classroom teachers through online video and multimedia resources through partnerships with national classroom teachers and the San Diego Supercomputer Center, University of California San Diego. Plans will be developed to provide online workshops ("web casts") to help teachers learn to use technology to enhance their teaching and support their students. The workshops will be designed to cover computer basics; web design and development; chemistry; geoinformatics; and bioinformatics.

Cyber-infrastructure refers to the ability to access and integrate today's hardware, software, and human information technology resources in order to facilitate science, engineering, research and education goals. In the next five years, the challenge for Cyber-infrastructure will be socio-cultural and behavioral dynamics, economics, and policy. The objective of this award is to hold a hold a workshop that will seek to identify for the SBE and CISE Directorates at NSF the key problems in the social sciences, economics, organizational, and policy studies for Cyber-infrastructure. The goal of the workshop will be to stimulate discussion within and between the communities served by the SBE and CISE Directorates about Cyber-infrastructure and to produce a report that outlines some of the key challenges and areas for fruitful research, development and experimentation by SBE and CISE. Specific goals for the workshop are
1) To produce a report which lays out a Cyber-infrastructure research and development roadmap for the SBE and CISE community and provide a framework for projects and efforts going forward;
2) To provide a venue for community building within the SBE and IT communities;
3) To lay out a program for research on the effects of IT on society and the
dynamics of IT-focused organizations and the cyber-culture

Broader Impacts:
SBE plays a key role in the development of Cyber-infrastructure. Not only is data and computational Cyber-infrastructure an enabler for key efforts for SBE communities, but SBE can play a critical role in the design, development and culture surrounding the deployment of Cyber-infrastructure itself. A strategic approach to the development of a body of research, development, and experimentation that addresses the socio-cultural, economic, and policy challenges of Cyber-infrastructure will be critical to ensure its success. In particular, social and behavioral scientists, humanists, organizational, policy and management researchers, etc. are critical as process builders for Cyber-infrastructure as well as end users of Cyber-infrastructure. Such communities have only sporadically been engaged in the development of Cyber-infrastructure models, social and organizational structures and have a wealth of experience and context to offer.

Cyberinfrastructure refers to the ability to access and integrate today's hardware, software, and human
information technology resources in order to facilitate science and engineering goals. This proposal
describes a plan of action for building and delivering cyberinfrastructure. The efforts of the San Diego Supercomputer Center (SDSC) will focus on the provisioning of a deeply integrated environment,
which coordinates hardware, software, and human resources to engage and support users, and the
development of critical infrastructure to drive research and education.

SDSC activities will have particular focus on building the data cyberinfrastructure so critical to
managing, understanding, and gaining knowledge from today's deluge of data. The Center will also focus on building the community for cyberinfrastructure: extending the reach of researchers and educators through deep collaborations, professional-level services and software, large-scale data management and
computational facilities, and the engagement and support of new communities of social scientists and
computer scientists who can play an important role in cyberinfrastructure as process-builders.

Cyberinfrastructure extends the reach of researchers and educators past the capabilities, challenges, and
limitations of their home environment.. Cyberinfrastructure provides a vehicle for accessing additional
infrastructure that benefits users significantly in their research and education efforts. SDSC will be a core
resource for cyberinfrastructure through highly valued, applied, and multi-disciplinary expertise and
community leadership; large-scale data and computational resources; and robust, professional, user-
focused software. SDSC will provide:

o A broad spectrum of cyberinfrastructure software and services that enable users to coordinate
available technological tools in their home environment with the larger landscape of cyberinfrastructure resources and impact their science efforts.

o A Synthesis Center in partnership with the California Institute for Telecommunications and
Information Technology (Cal-IT2) [1] for leveraging infrastructure across projects and disciplines.
o Intensive strategic collaboration projects among domain scientists and computer scientists that
bridge science and technology goals.

o An Advanced Cyberinfrastructure Laboratory that will be a flexible environment to support research, development, experimentation, and evaluation with cyberinfrastructure technology and tools.

o A comprehensive Integrated Data Cyberinfrastructure to serve a broad class of existing and emerging data-oriented users.

The customers and beneficiaries of cyberinfrastructure are its users, and the ultimate metric of success
for cyberinfrastructure is the ability to focus on the challenges of the science rather than the challenges of
the tools in driving new discovery. SDSC will build and deliver key services, infrastructure, and
innovations that will enhance the ultimate success of cyberinfrastructure. Together with a community of
collaborators including the National Center for Supercomputing Applications (NCSA), TeraGrid/ETF,
project-focused community efforts, and a broad group of domain scientists, engineers and computer
scientists, SDSC will provide a cohesive combination of expertise, hardware, and software to facilitate the
transformation of the cyberinfrastructure vision into reality.

Increasingly, intellectual content is 'born digital', as a consequence of which the digital library and digital archival communities find themselves faced with unique preservation challenges, challenges that call for comprehensive digital preservation lifecycle management processes. These processes, in turn, need to be unobtrusive, with minimal impact on content production and dissemination. In order to make it as easy as possible for content creators to preserve their content for the long-term, preservation processes should be integrated into the content production lifecycle. This project will demonstratef a software technology that will comprise such a preservation environment for a video collection that includes other related multi-media content such as audio, transcripts, annotations, related and introductory Web pages, and descriptive, technical, and rights metadata currently being captured in a FileMakerProTM database. The proposal is to take an existing video production workflow that is currently being used for producing and Web-casting video content in a small-scale studio setting, and to integrate it with a digital preservation life-cycle management process that will enable the digital content to be archived for long-term preservation. The collection, 'Conversations with History', is produced at the University of California, Berkeley, edited by University of California, San Diego TV (UCSD-TV), and broadcast and Web-cast through UCTV.

The major research and development challenge to address the development and automation of preservation processes that comprehend the accession, description, organization, and preservation of the video collection and associated content. The automation of these processes requires the integration of workflow systems that are used in active production with preservation systems, and the tracking of the execution of the preservation processes for completion and error recovery. The system will demonstrate an effective preservation methodology by 1) ensuring that preservation 'of last resort' is available for at-risk multi-media collections of critical lasting importance; 2) ensuring that preservation is available for collections that will expand and evolve over time, while managing evolution in the underlying software technologies; and 3) demonstrating a standard reference model for digital preservation lifecycle management that can be integrated into active production workflows.

INTELLECTUAL MERIT: A report from the April 2002 workshop sponsored by the NSF and the Library of Congress states 'Solutions are urgently needed to prevent further loss of valuable digital information - these problems are urgent - action is needed now, not some time in the future'. The proposed demonstration of digital preservation lifecycle management addresses this critical problem by showing that current software systems provide the required functionality, can be implemented at scale, and can preserve at-risk collections. The demonstration will show an ingestion model and collection management for multi-media material, in which preserved intellectual capital is easily accessible, and research results, educational material, and new knowledge can be incorporated smoothly over the long term. The system integrates digital library, data grid, and persistent archive technologies, with the goal of providing a fundamental building block for long-term collection management and preservation.

BROADER IMPACT: The NSF Cyberinfrastructure program, the Library of Congress National Digital Information Infrastructure and Preservation Program (NDIIPP), and the National Archives and Records Administration (NARA) Electronic Records Archive all are actively concerned with the preservation of data. The demonstration of the ability to extract material from a production environment and import the
material into a scalable preservation environment is essential for making preservation viable. The demonstration of publication, preservation, federation, and analysis will serve as a proof of preservation
concepts to a broad community including humanists, social scientists, scientists, engineers, students, and
the general public. The extension of the demonstration project into a production facility will enable other
public and at-risk digital content to be preserved. The preservation architecture will support federation of
existing digital libraries and preservation environments, enabling the expansion of the facility to all
communities. It will provide a replicable model that will demonstrate a system that can contribute to a
scalable prototype for a truly national preservation backbone.

OCI 0737721
Francine Berman
UCSD
"Blue Ribbon Panel on Sustainable Digital Preservation and Access"


The Blue Ribbon Panel on Sustainable Digital Preservation will seek to clearly and compellingly outline the problem, analyze relevant work and best practices, and provide actionable recommendations to support the development of an economically viable and effective national framework for digital preservation. The group plans to work closely with the NSF, the research, education and preservation communities, and other interested agencies and foundations (including the Library of Congress, the Mellon Foundation, the UK Joint Information Systems Committee, and others) to achieve these goals, and to maximize the impact of the panel's proceedings, produce effective and useful written materials, and deliver a compelling Final Report which focuses attention on this critical grand challenge and helps illuminate a way forward.

The broader impact of this workshop is as follows:
The Workshop noted "a critical need for research and development of economic and business models that will sustain digital preservation ... New business models are needed to make digital preservation affordable to individuals, government agencies, universities, cultural institutions, and society at large."

The intellectual merit of the workshop is that the need for such infrastructure will only become more acute as our scientific knowledge and cultural record in short, the information based raw materials with which we discover, learn, and prosper, along with the environments in which we produce and consume information continue to migrate to the digital realm.