ARL PET Project

ARL-CY5-IC--1

Institution Name: Florida State University

Project Identifier: ARL-CY5-IC--1

Project Title: I/C Core Support

POC: Fox, Geoffrey C

Email: gcf@cs.fsu.edu

Phone: 850 644 4587

Fax: 850 644 0098

CTA: IC

Project Description: Information and communications technologies are areas of both
tremendous interest and rapid change. These technologies can be used
within the DoD HPCMP for a wide range of purposes, including:
o Making training and education more accessible
o Facilitating collaboration of geographically distant researchers
o Simplifying dissemination of information to interested parties
o Simplifying access to and use of HPC systems
o Making it easier to produce and manage large, complex computational
simulations
o Providing more convenient approaches to the storage, presentation,
and analysis of large data sets

The I/C Core project provides resources to allow the Florida State
University team to track and assess these technologies as they evolve,
providing a basis for the transfer of sufficiently mature technologies
into the HPCMP domain, and for Focused Efforts aimed at developing a
more detailed understanding of these techologies and exploring their
application in areas relevant to the HPCMP. I/C Core also provides the
basis our user outreach efforts designed to bring these technologies
and applications to the attention of those who can benefit from them
within the HPCMP. As a cross-cutting support area, our primary
outreach interactions are with others on the PET team supporting
specific CTAs and other cross-cutting areas who in turn can convey the
information to users who may benefit. We also reach out directly to
users through training activities (covered in a separate proposal),
publications and presentations in many forums.

The past year has been an eventful one in the I/C area, and it has
helped to refine and to some extent reorient our areas of emphasis
relative to previous years. Here are some of the areas on which we
plan to focus our technology tracking and assessment activities this
year...

Portals for Computational Science

In the last year, the "portal" has emerged as the dominant concept for
the construction of Internet applications and services. Areas of
particular interest to the HPCMP include portals which facilitate
access to HPC resources and software and the collection, storage, and
analysis of large volumes of data. Portals for training and education
are also important to the PET mission.

On the computing portals side, we will continue our involvement in the
Grid Forum and Computing Portals Forum, which seek to develop
standards for interoperability of computing portal systems. These
forums, along with other conferences also play an important role in
tracking the wide variety of efforts currently active in this area.
An important new area of development in the last year has been the
emergence of a second generation of frameworks for the building of
portals. These systems, such as e-Speak from Hewlett-Packard and
Ninja from Berkeley, are built on top of CORBA, one of the main
first-generation frameworks for the construction of portal middleware,
and claim to allow much faster and more efficient portal development.

Related to computing portals are those which support the collection,
storage, and analysis of large volumes of data, as is often required
in the T&E community. While the simple provision of web-linked
databases to provide access to large data stores is by now fairly
straightforward in concept, there remain many open questions as to how
best to integrate this capability with the kind of analytical and
predictive tools that should be applied to the data, and how to
provide the level of flexibility required to satisfy T&E needs (most
web-linked database work involves fixed pre-defined queries). We will
work closely with the PET IMT team to help provide information on the
best available technologies for these purposes and how they can be
integrated with other portal-building activities.

Training and Education Technology

The portal concept is important not only in the context of scientific
computing, but also for training and education. There are an
increasing number of both academic and commercial technologies/systems
becoming available in this arena. We plan to look more closely at a
number of these systems, such as IBM Techxplorer, LearningSpace,
Blackboard, WebCT. The shift to Florida State where Fox has a State
supported research effort in distance education brings new resources
-- especially understanding of asynchronous tools Blackboard and
Web-MC. We propose to work closely with the on-site Training team to
understand the needs and use of training technology at the MSRC in the
context of the changing technology base.

At the same time, education is fundamentally a form of collaboration,
and for delivery in a network environment, collaboration technologies
also play a crucial role. For this reason, we see additional focused
efforts are likely to arise from the collaboration whitepaper discussed
below.

Collaboration Technologies

Although network-based collaboration technologies have been around for
some time, and the HPCMP has made extensive use of the "m-bone" and
Tango Interactive systems in selected application, electronic
collaboration tools have not captured mainstream interest. More
recently thanks in part to improvements in network connectivity, an
increasing number of commercial entrants into the collaboration area
(in particular the voice-over-IP Internet telephony boom), and new
web-based standards (i.e. SMIL), interest is starting to pick up. We
will track the evolution of these approaches and provide assessments
of selected systems, primarily in the context of collaborative portals
for education and computational science.

At Dr. Radha's request, we will provide leadership on the development
of a PET whitepaper with ideas on how to foster both administrative
and technical collaboration within the HPCMP. This whitepaper is
expected to provide the basis for an ARL MSRC initiative in
collaboration, which will include additional PET projects we will
assist in developing and implementing.

At the technological "high-end" of collaboration environments, the
"Access Grid", developed at Argonne National Laboratory and adopted by
NCSA's Alliance PACI program. Access Grid is a general specification
of hardware capabilities which provides sophisticated audio and video
capabilities and can be used with a wide range of software for both
collaboration and remote visualization. We have recommended
installing Access Grid-compatible hardware as part of the effort to
provide better links between the APG and Adelphi sites of ARL, and we
will provide support and assistance with integration of this
technology (which we also plan to install at Florida State).

Project Objectives: Florida State University will provide a core level of effort to
support technology tracking and transfer, user outreach, assessment of
tools and technologies, and long-range leadership to facilitate use of
the best available information and communication technologies by the
PET team, the ARL MSRC, and its users.

Deliverables: 1) Contributions to PET reports and presentations at review meetings,
as required
2) Leadership of the collaboration whitepaper and advising on the ARL
PET collaboration initiative
3) Participation in national and international meetings, conferences,
and other forums. Trip reports to be delivered as appropriate.
4) Selected experimental deployment and assessment of identified
collaboration and education tools/technologies

Customers/End Users: This work helps insure that the entire PET team stays current with
relevant I/C technologies, which are then incorporated into PET
projects, trainings, and disseminated to HPCMP users in other ways.

Benefit to Warfighter: Through these core activities, we help insure that HPCMP researchers
are in a position to take advantage of appropriate
commodity/commercial technologies in the information and
communications technology arena

Project Dependencies: This proposal assumes support from Derek Moses, part-time I/C on-site
lead, who is not an FSU employee. It also requires interactions with
the on-site Training team and the IMT team in the form of discussions
and consultations.

Much of the technology tracking occuring under this project directly
influences our Focused Efforts, providing the kind of input required
to insure that the Focused Efforts remain at the forefront of
technology.

We have identified two mini-projects, in the Training Technologies
area:
1) Assessment and deployment of RealNetworks tools for integrated
multimedia training materials, and
2) Assessment of new standards for training material (IMS from
Educause and SCORM from DoD's Advanced Distributed Learning project)
and their impact on the PET Training program, and experimental
deployment of standard-compliant adaptations of training materials

These projects go somewhat beyond typical "Core Support" activities,
but are not terribly large (we estimate about $20k each). Therefore
we plan to integrate these mini-projects into forthcoming proposals on
Training and on Collaboration. However if the management prefers
another approach, we can incorporate them into Core or cast them as
standalone proposals.

Risk Element: As this project primarily involves tracking technologies, this project
itself has minimal risk, and helps to reduce the risk of technological
irrelevance or obsolescence in other PET activities. However, as in
any fast-moving field, there is a risk that the importance of "stealth
technologies" may not be recognized initially.

Institution Name:	Florida State University
Project Identifier:	ARL-CY5-IC--1
Project Title:	I/C Core Support
POC:	Fox, Geoffrey C
Email:	gcf@cs.fsu.edu
Phone:	850 644 4587
Fax:	850 644 0098
CTA:	IC
Project Description:	Information and communications technologies are areas of both tremendous interest and rapid change. These technologies can be used within the DoD HPCMP for a wide range of purposes, including: o Making training and education more accessible o Facilitating collaboration of geographically distant researchers o Simplifying dissemination of information to interested parties o Simplifying access to and use of HPC systems o Making it easier to produce and manage large, complex computational simulations o Providing more convenient approaches to the storage, presentation, and analysis of large data sets The I/C Core project provides resources to allow the Florida State University team to track and assess these technologies as they evolve, providing a basis for the transfer of sufficiently mature technologies into the HPCMP domain, and for Focused Efforts aimed at developing a more detailed understanding of these techologies and exploring their application in areas relevant to the HPCMP. I/C Core also provides the basis our user outreach efforts designed to bring these technologies and applications to the attention of those who can benefit from them within the HPCMP. As a cross-cutting support area, our primary outreach interactions are with others on the PET team supporting specific CTAs and other cross-cutting areas who in turn can convey the information to users who may benefit. We also reach out directly to users through training activities (covered in a separate proposal), publications and presentations in many forums. The past year has been an eventful one in the I/C area, and it has helped to refine and to some extent reorient our areas of emphasis relative to previous years. Here are some of the areas on which we plan to focus our technology tracking and assessment activities this year... Portals for Computational Science In the last year, the "portal" has emerged as the dominant concept for the construction of Internet applications and services. Areas of particular interest to the HPCMP include portals which facilitate access to HPC resources and software and the collection, storage, and analysis of large volumes of data. Portals for training and education are also important to the PET mission. On the computing portals side, we will continue our involvement in the Grid Forum and Computing Portals Forum, which seek to develop standards for interoperability of computing portal systems. These forums, along with other conferences also play an important role in tracking the wide variety of efforts currently active in this area. An important new area of development in the last year has been the emergence of a second generation of frameworks for the building of portals. These systems, such as e-Speak from Hewlett-Packard and Ninja from Berkeley, are built on top of CORBA, one of the main first-generation frameworks for the construction of portal middleware, and claim to allow much faster and more efficient portal development. Related to computing portals are those which support the collection, storage, and analysis of large volumes of data, as is often required in the T&E community. While the simple provision of web-linked databases to provide access to large data stores is by now fairly straightforward in concept, there remain many open questions as to how best to integrate this capability with the kind of analytical and predictive tools that should be applied to the data, and how to provide the level of flexibility required to satisfy T&E needs (most web-linked database work involves fixed pre-defined queries). We will work closely with the PET IMT team to help provide information on the best available technologies for these purposes and how they can be integrated with other portal-building activities. Training and Education Technology The portal concept is important not only in the context of scientific computing, but also for training and education. There are an increasing number of both academic and commercial technologies/systems becoming available in this arena. We plan to look more closely at a number of these systems, such as IBM Techxplorer, LearningSpace, Blackboard, WebCT. The shift to Florida State where Fox has a State supported research effort in distance education brings new resources -- especially understanding of asynchronous tools Blackboard and Web-MC. We propose to work closely with the on-site Training team to understand the needs and use of training technology at the MSRC in the context of the changing technology base. At the same time, education is fundamentally a form of collaboration, and for delivery in a network environment, collaboration technologies also play a crucial role. For this reason, we see additional focused efforts are likely to arise from the collaboration whitepaper discussed below. Collaboration Technologies Although network-based collaboration technologies have been around for some time, and the HPCMP has made extensive use of the "m-bone" and Tango Interactive systems in selected application, electronic collaboration tools have not captured mainstream interest. More recently thanks in part to improvements in network connectivity, an increasing number of commercial entrants into the collaboration area (in particular the voice-over-IP Internet telephony boom), and new web-based standards (i.e. SMIL), interest is starting to pick up. We will track the evolution of these approaches and provide assessments of selected systems, primarily in the context of collaborative portals for education and computational science. At Dr. Radha's request, we will provide leadership on the development of a PET whitepaper with ideas on how to foster both administrative and technical collaboration within the HPCMP. This whitepaper is expected to provide the basis for an ARL MSRC initiative in collaboration, which will include additional PET projects we will assist in developing and implementing. At the technological "high-end" of collaboration environments, the "Access Grid", developed at Argonne National Laboratory and adopted by NCSA's Alliance PACI program. Access Grid is a general specification of hardware capabilities which provides sophisticated audio and video capabilities and can be used with a wide range of software for both collaboration and remote visualization. We have recommended installing Access Grid-compatible hardware as part of the effort to provide better links between the APG and Adelphi sites of ARL, and we will provide support and assistance with integration of this technology (which we also plan to install at Florida State).
Project Objectives:	Florida State University will provide a core level of effort to support technology tracking and transfer, user outreach, assessment of tools and technologies, and long-range leadership to facilitate use of the best available information and communication technologies by the PET team, the ARL MSRC, and its users.
Deliverables:	1) Contributions to PET reports and presentations at review meetings, as required 2) Leadership of the collaboration whitepaper and advising on the ARL PET collaboration initiative 3) Participation in national and international meetings, conferences, and other forums. Trip reports to be delivered as appropriate. 4) Selected experimental deployment and assessment of identified collaboration and education tools/technologies
Customers/End Users:	This work helps insure that the entire PET team stays current with relevant I/C technologies, which are then incorporated into PET projects, trainings, and disseminated to HPCMP users in other ways.
Benefit to Warfighter:	Through these core activities, we help insure that HPCMP researchers are in a position to take advantage of appropriate commodity/commercial technologies in the information and communications technology arena
Project Dependencies:	This proposal assumes support from Derek Moses, part-time I/C on-site lead, who is not an FSU employee. It also requires interactions with the on-site Training team and the IMT team in the form of discussions and consultations. Much of the technology tracking occuring under this project directly influences our Focused Efforts, providing the kind of input required to insure that the Focused Efforts remain at the forefront of technology. We have identified two mini-projects, in the Training Technologies area: 1) Assessment and deployment of RealNetworks tools for integrated multimedia training materials, and 2) Assessment of new standards for training material (IMS from Educause and SCORM from DoD's Advanced Distributed Learning project) and their impact on the PET Training program, and experimental deployment of standard-compliant adaptations of training materials These projects go somewhat beyond typical "Core Support" activities, but are not terribly large (we estimate about $20k each). Therefore we plan to integrate these mini-projects into forthcoming proposals on Training and on Collaboration. However if the management prefers another approach, we can incorporate them into Core or cast them as standalone proposals.
Risk Element:	As this project primarily involves tracking technologies, this project itself has minimal risk, and helps to reduce the risk of technological irrelevance or obsolescence in other PET activities. However, as in any fast-moving field, there is a risk that the importance of "stealth technologies" may not be recognized initially.