Representative FY 2002 agency activities
NSF: Continue exploring quantum phase data storage and
retrieval; shared-memory multiprocessor design; nanoscale device and
system architectures
DARPA/NSF: Investigate the use of DNA-like molecules to store and
compute terabyte-scale problems
DARPA: Investigate very large- scale integration of photonics for
intra- and inter-chip communication, including processor-in-memory
arrays
NSA: Continue research to demonstrate the feasibility of quantum computing
devices and other high-performance, superconducting alternatives to
current silicon and gallium arsenide technologies; architecture, configuration,
and programming of "smart memory" chips
NIST: Research in quantum computing, secure quantum communication,
photonics, nanotechnologies, optoelectronics, and new chip designs
and fabrication techniques
ODUSD (S&T): University-based research in quantum communications
and quantum memory
|
The complementary metal oxide semiconductor (CMOS) chip,
the two-dimensional miniature electronic map on a silicon wafer that
is the standard building block of computing systems, is fast approaching
its physical limits. That is, the electronic signals that can be routed
through its pathways are finite in quantity and speed. Even the complex
technical amassing of chips that produced Option White demonstrates
these limitations in the machine's enormous size and power requirements.
The NITRD agencies are examining new materials and methods to create
wholly new designs for processors in computing devices. Federal missions
and private-sector IT innovation alike require both mid-term incremental
improvements in computational speeds and long-term breakthroughs to
radically new processor architectures capable of teraops and petaops
speeds.
In FY 2002, the agencies will support research at the theoretical
and empirical intersection of biology, information science, and micro-electromechanical
systems [Bio:Info:Micro]. Advances in photonics, nanotechnologies,
sensors, actuators, optoelectronics, digital, analog, and mixed signal
processing, and new fabrication technologies make it possible, for
example, to conceive of integrated designs in 3-D on a chip with billions
of transistors. This work focuses on designing new, modular hybrid
architectures that include fault-tolerance, programmability (including
novel approaches such as amorphous computing methods), and security
features needed in embedded systems for defense.
A related research area is biological substrate computing, the potential
in organic molecules - such as deoxyribonucleic acid (DNA), ribonucleic
acid (RNA), and proteins - to provide vast storage and processing
capacities. For example, one gram of DNA contains 1021 DNA bases,
which is equal to 108 terabytes of information storage. Breakthroughs
in this area could result in:
- High-volume, content-addressable storage
- Solutions to computationally hard problems that are not
now solvable
- Self-assembly of nanostructures using DNA/RNA tiling.
The
nanostructures in turn could be used for nanoscience such
as molectronics (described below)
|
Because the timing of revolutionary breakthroughs in all these fundamental
investigations cannot be predicted, the NITRD agencies also continue
to explore superconducting materials and related technologies that
offer the potential to produce the substantial incremental improvements
in processing speeds that will be needed in the near- and mid-term.
Agencies will support long-range research to explore the potential
of atomic phenomena - such as quanta of light or molecular nuclei
- to serve as high-speed processing mechanisms. This area holds great
promise as a future means of providing:
- Ultrasecure communications over optical backbone networks
- Orders of magnitude increases in the speed of algorithms
such as for searching unsorted databases or factoring large
numbers
- Quantum computers that can give detailed and faithful
simulations of molecular processes and phenomena in physics
|
In addition, agencies will support research to attain long-term breakthroughs
in computer design and fabrication. The potential results of these
efforts include:
- Innovative computational structures, 3-D architectures,
hybrid technologies
- Reconfigurable systems on a chip, adaptive and polymorphous
computing
- Processor in memory (PIM) and other efforts to provide
memory performance commensurate with processor performance
- New computational substrates:
|
- Quantum computing
- Biological substrate computing
- "Smart fabric." Using technology for interweaving
battery, fiberoptic cable, and metal connectors, scientists
can produce fabric that can be embossed with enough processors
to provide on-person processing on the order of tens of
teraops (the size of today's larger supercomputers)
- Molectronics: computation at the molecular scale, which
holds the potential of providing extremely fast, high-density
processing power for the next generation of strategic
computing for the military
|
|
