This is historical material, "frozen in time."
The web site is no longer updated and links to external web sites and some internal pages will not work.
Appendix A: Scientific and Technical
Goals and Objectives for Information and
Communications R & D
The Information and Communications (I&C) technology system is complex and dynamic.
New technology flows from "developers" to "users" along increasing stages of maturity
(i.e., from conceptual, early-stage science, to experimental mid-stage technology
development, to mature, late-stage infrastructures in support of the next generation of
science and technology). Technology also flows along increasingly complex and abstract
levels of system structure (e.g., from primitive components assembled into
communications systems, which are assembled into computing systems, and on through
Although it is easy to describe this flow of new technology as a linear process, reality is
more complex. New knowledge and technologies develop in parallel and mix among the
activities, ultimately progressing toward the goals set forth by the Strategic Focus Areas.
Maturing technologies reach a stage where they no longer represent research in their own
right. At this point they not only enter commercial development and broad scientific
deployment, but they also play a critical role in enabling research and development.
Examples include mature generations of scalable parallel multiprocessors supporting the
Grand Challenges of science and engineering and moderate speed, and wide-area
networks connecting "virtual" user communities to distributed resources. These become
key elements of an underlying infrastructure for research and development.
Often the limitations of existing infrastructures encountered in their operational use (as
they are scaled up or used in more diverse ways) motivates new conceptual
breakthroughs. The development of the next generation of innovative technologies is
fueled by providing developers with access to the latest infrastructures, thus providing a
context in which to develop alternatives that address the fundamental limitations of
As noted in Section 2.3, CIC has developed a taxonomy that characterizes the technology
base underlying I&C research and development into seven broad areas. These areas
represent the major building blocks of Agency-specific Federal Programs:
Components are the primitive physical building blocks of communications,
computation, and information processing systems;
Communications addresses the technologies, services, and infrastructure of
interconnected systems exchanging information in analog and digital forms;
Computing Systems focus on the structure, organization, and design of hardware and
software for computational and information processing systems;
Support Software and Tools are the technologies for developing and managing the
software creation, evolution, and maintenance processes, including applications
building blocks and the operational software needed to provide system-level services
Intelligent Systems develop techniques for automated reasoning to advance the
capabilities for aids to human-decision making and execution;
Information Management develops the capabilities for information retrieval, data
management, and archival storage in support of applications;
Applications of specific interest to the CIC are those that use information technology
in innovative ways, with intensive computation, information, communication, or
As part of the process to develop this strategic implementation plan, an informal budget
crosscut was conducted to assess the overall distribution of CIC resources in the seven
technology R&D activity areas, as well as the proportionate share currently invested in
the Federal HPCC Program. These are reported in Figure 2 above. Information and
Communications R & D funding is currently approximately $2 billion per year, of which
half is for the HPCC Program. Since this represents the R&D investments, the
government leverage of its investment to the entire worldwide industry is only 250:1.
It is important that these numbers should not be considered definitive crosscuts, and it is
not CIC's intention to recommend crosscuts by the strategic focus areas at this time.
Rather, CIC intends to use this important baseline, coupled to detailed implementation
plans, such as the Federal HPCC Program and Agency submissions, to evolve a
meaningful, streamlined approach based on the strategic focus areas to understand and
manage important cooperative efforts in the future.
The following sections elaborate on the goals, opportunities, ongoing activities, new
efforts and challenges for each of the seven I&C broad areas of research.
The goal of the Components area is to develop component technologies over the next
decade that will drive advances in computing and information technology. Component
technology not only provides the information processing devices at the heart of
computing and communication equipment, but also the interfaces with which information
appliances interact with the physical world and with people. This effort will build on the
National Electronics Manufacturing Initiative (NEMI) of the CIT committee.
Cheaper, smaller, and faster electronic components will improve today's applications and
enable information technology to be applied to new domains, perform new services, and
create new markets in the future. Revolutionary improvements in "bits per second per
dollar" will be required to meet the demand of video-oriented customers for increasingly
individualistic services that draw on global information. Visual communication requires
improved displays and imagers. To balance increased processing and communication
capability, mass data storage technology must continually improve, which demands the
mastery of complex low-cost electromechanical systems. Improved radio frequency (RF)
and infrared (IR) components are needed to support wireless connectivity, which implies
that digital systems with lower power requirements must be developed, and that energy
storage subsystems must improve. Interfacing these systems effectively with the physical
world will require low cost sensors, sources, actuators, and displays, often customized to
the application. Product development times and lifetimes continue to decrease as capital
equipment costs explode. Modeling and simulation are critical to shortening the
development cycle and preventing costly mistakes. Future opportunities lie in such areas
as: optoelectronic components (with their enormous intrinsic bandwidths and low-loss
fiber compatibility); the merging of microelectronics and micromechanics; and the newly
emerging area of molecular electronics.
Present R&D is focused on: processes for electronic components; optoelectronics; mass
data storage; energy storage; electromechanical systems; modules and processes for
interconnect systems; and electromechanical assembly, including packages, boards, and
products. The driving force in semiconductor design and manufacturing is increasing
miniaturization and function integration. In the area of optoelectronic displays, activities
include: improved cathode ray tubes; flat panel and head-mounted displays using active
matrix liquid crystals, electroluminescence, plasma, and cold cathode technologies;
projection displays using digital micromirrors, liquid crystals, and laser projection.
New efforts are required in advancing display processing technology, such as high speed
video processor modules, workstations, high bandwidth busses, image transmission over
packet networks, and high bandwidth, digitally compressed video and data systems. New
compression algorithms, graphics tools, graphics standards, and user interfaces also must
Infrastructure broadly applicable to component design and manufacturing is needed that
will address such areas as: design tools; design for "X" (manufacturability, reliability,
etc.); manufacturing automation; testbeds, including generic equipment and software;
rapid prototyping and related services, including electronic and optoelectronic
Longer term component research needs include: nanotechnology; complex physical
device simulation; biology-based computing components (including sensors);
fundamentals of electronic materials; optical components and systems; CAD synthesis
algorithms to reduce time to market; models and characterization of devices,
interconnects and packaging; methods for integrating processing into components such as
sensors, displays, and mass storage devices; testbeds for evaluating or comparing
component designs; test theory and algorithms; components for analog information
processing. As components take on more function and become more complicated,
techniques must be developed that can convert high-level descriptions of their
requirements and specifications into efficient physical designs. Components using
increasingly disparate technologies in complex combinations cannot be designed from
theory alone. For device characterization, prototypes of these components must be built
and measured to determine their behavioral characteristics in practice. As levels of
integration increase, entire subsystems may be reduced to single components. Testbeds
for comparing prototypes of such devices in realistic application environments will be
The goal of the Communications area is to provide the necessary communications
infrastructure and technologies to support the advanced requirements of the National and
Grand Challenge class applications as well as to address the ubiquitous and universal
accessibility required by the citizenry, industry, academia, and government sectors.
The successful deployment and evolution of the NII requires communications and
information support services that are secure, reliable, scalable, fault-tolerant, easy-to-use,
easy-to-manage, affordable, and deployed across a wide variety of technologies and
media. Such systems will enable more people and businesses to use the NII for both
personal and business advantages while building the critical mass necessary for moving
to a post-industrial information-based society. In order to provide such a capability,
research is particularly important in those areas that make it easy for the end-user to
access the NII and its contents in a secure and private manner, that ensure the
interconnectivity and interoperability of various technologies and services in a seamless
fashion, that support real-time, distance-insensitive interactions and collaborations, and
that support nomadic and local independent access.
Advanced communications theory and algorithm research already initiated that must be
pursued includes: analog and digital wireless services, micro to macro cellular systems,
scalable, reliable, adaptive routing algorithms that support mobility and enhanced
multicast capabilities, inter- and intra-enterprise security mechanisms, network resource
allocation and management, all-optical switching and transmission techniques, wireless
and satellite technologies, nomadic computing and access, internetworking and
interoperable services, encoding and compression techniques, and support for electronic
Areas that require new or enhanced efforts include: universal accessibility, national
digital wireless coverage, autonomous self-configuring network technology (plug and
play), 100 gigabits per second network capability available generally, and terabits per
second network capability available in the lab.
Many challenges remain in the areas of: all optical switching; ubiquitous, multi-domain,
and multi-level security; evolution of legacy systems and applications; global
interoperability; ease-of-use; and NII user affordability, identification, and privacy.
A.3. Computing Systems
The goal of the Computing Systems area is to develop and demonstrate advanced
computer systems and architectures offering very high performance, very low cost for a
given level of performance, and very high performance in a small sized footprint, for a
broad range of scientific, engineering, industrial, and national security applications.
A unifying opportunity for this area is the development and demonstration of balanced
parallel systems that can gracefully scale across a wide range of underlying numbers of
processor nodes and interconnection structures. The major issues in these developments
are: insertion of ever more powerful processing nodes, incorporating hierarchical degrees
of parallelism; faster interprocessor communications; global management of memory and
data in cooperation with systems software; scalable, parallel input/output (currently a
major bottleneck in recovering and interpreting the massive amounts of data produced by
most high performance applications); and architectural convergence towards scalable
parallel processing with support for vector processing, multiprocessing, message passing,
and shared memory abstractions in a unified architecture.
Activities contributing to balanced, scalable, parallel systems include more powerful
processing nodes, faster interprocessor communications, global management of memory
and data, scalable parallel input/output subsystems, and support for vector processing,
multiprocessing, message passing, and shared memory abstractions in a unified
Architectural alternatives and processor, memory, interconnect, and I/O technology to
support them are also being investigated and include scalable parallel systems and
multicomputers, homogeneous and heterogeneous local-area clusters of commodity
workstations, and distributed processing systems.
These advancements depend upon basic research that includes developing a fundamental
understanding of architectural design, interconnection structures, computational
complexity, and models of computation applied to parallel and distributed algorithms and
systems. The theoretical underpinnings of computer-based aids for hardware and system
design, such as reliability analysis, graph theory, and system timing methodologies,
contribute to the science of computing system design.
New research and development is required that is focused on more specialized systems
for supporting ubiquitous access, fault-tolerant highly available computing, real-time
response in industrial and control applications, highly portable systems that can compute
and communicate on the move, and other application-specific processors for tasks like
signal processing. System support functions, such as security, reliability, resource
scheduling, load balancing, performance instrumentation, and visualization are also being
developed in support of this effort.
The key challenge is to provide the technology to enable the low cost, ubiquitous systems
required by user-centered interfaces and tools. One must be able to configure scalable
systems over a wide performance range, spanning from workstation-class machines to
networks of workstations to large scale parallel computing systems capable of
teraoperations per second performance and beyond. System and application software that
can gracefully scale across this range is also a key piece of this challenge.