Getting Us Over the Brick Wall
Rita Glover, EDA Today, L.C.
April 2003
Technology has progressed to the point that materials can be
manipulated at the level of their individual atoms. In the coming
decades, this nanoscale technology will have pervasive and profound effects
on the electronics industry.
As electronic devices become much smaller and denser, new
architectures and mechanisms for information transport must emerge to meet
the challenges of increasing complexity and power dissipation. Just as
the electronics industry will change significantly with the introduction and
evolution of nanotechnology, so too will the work of electronic engineers.
To inform working engineers on these revolutionary changes,
a NanoEngineering Education Workshop was held in conjunction with DesignCon
2003 in late January (Santa Clara, CA) to assemble technologists from
government, industry, and academia who are at the forefront of this exciting
field. We talked with Barry Sullivan, Ph.D., Director of Content
Development for the International Engineering Consortium, Chicago, Illinois,
who illuminated the distinctions between nanoscience, nanotechnology, and
nanocomputing.
Science is about the discovery and observation of things as
they exist in nature. Technology is the realm of the engineer,
involved in the process of taking these scientifically discovered things and
manipulating, shaping, and applying them. We have been doing
nanoscience for decades, through work at the molecular level at the scale of
nanometers. But when an engineer starts manipulating materials at the
atomic level, it's nanotechnology.
A distinction needs to be drawn between nanotechnology and
simply "doing small things." Chips have shrunk to 90 nanometer
processes just by following the normal progression of Moore's Law into ever
smaller geometries. But the term nanotechnology applies only if an
engineer is actually manipulating the materials on an atomic level.
"Quantum computing" is an interesting example of just how
revolutionary nanocomputing can be. Here, researchers are looking at
molecular-scale devices and even creating new kinds of devices that go
beyond the transistor.
In quantum computing, rather than transferring data based on
the movement of electronic charge, the data may be transferred based on the
spin states of atoms — their quantum state. If we can measure and
translate that state from one atom to the next, it could serve as a way of
propagating information.
This is something quite different from the transfer of
charge, or the movement of electrons. It's quite possible that
eventually the Van Neumann architecture may be implemented using some means
other than transistor-based logic.
Only a few years off, semiconductor development problems are
foreseen for which no solutions have yet been identified. It's clear
from the International Technology Roadmap for Semiconductors (ITRS) that a
brick wall is looming before us. We are certain to run into basic
physical limits if we continue doing things the way we have been, by just
scaling them down (or scaling them up, as the case may be).
Physics dictates that we can't just continue to do the same
things beyond a certain point. We either have to discover new physics
to get us past this point, or we have to come up with a completely different
way to address the issues.
The International Technology Roadmap for Semiconductors
(ITRS) indicates that current semiconductor technology will hit a brick wall
by 2008.
Source: Intel Corp.
EDA toolmakers now have solutions that work at the 90
nanometer node, but as devices become ever smaller and more transistors can
be etched into a given area, the complexity of designs becomes prohibitive.
If we continue on the current track, in five years the amount of heat
dissipated on a chip will be equivalent to the surface temperature of the
sun. Clearly, complexity and power dissipation are big parts of the
brick wall.
Nanotechnology offers an approach that could get us over the
brick wall. Here we're talking about doing things in a very different
way. Today, integrated circuits are created by basically carving away
at a piece of silicon to remove layers and etch out the desired structures.
Nanoelectronics, however, could bring us the ability to organically grow
these devices molecule by molecule, through a "self-assembly" process that
is additive rather than subtractive.
Nanoelectronics represents a true paradigm shift, because we
will truly be looking at a whole different way of design and fabrication.
Contributions to the solutions will probably come out of a combination of
academia, research institutions, as well as design tool developers.
Students are expressing interest in nanotechnology, and
academics are discussing how to initiate nanotechnology education.
Some institutions have started offering courses, often at the graduate
level. A report will soon be available from Electrical and Computer
Engineering Department Heads Association (ECEDHA) on schools where these
courses can be found.
For EEs who want to be at the cutting edge, now is the time
to start looking at nanotechnology in other domains, such as biomedical
engineering. Take note of the new ways in which those applications are
being fabricated, as well as the types of metrology being used to measure
their properties and control their manufacturing processes. This kind
of cross-disciplinary migration of knowledge will be key to the development
of the new field of nanoelectronics.
The NanoEngineering TecForum was jointly sponsored by the
International Engineering Consortium (www.iec.org)
and the Electrical and Computer Engineering Department Heads Association (www.ecedha.org).
For further information on the applications of nanotechnology, refer to the
Foresight Institute at www.foresight.org.
Top
