November 1, 1999
Computer Scientists Are Poised for Revolution on a Tiny Scale
SAN FRANCISCO -- Scientists at a variety of elite
laboratories around
the country are sharing a growing sense that they are
on the brink of a
new era in digital electronics. It will usher in a world of
circuits no more than
a few atoms wide, with a potential impact on computing, in
terms of speed
and memory, that may be too profound to fathom.
It was only in July that a group of researchers at
Hewlett-Packard and the
University of California at Los Angeles reported that they
had successfully
fashioned rudimentary electronic logic gates -- a basic
component of
computing -- that were the thickness of a single molecule.
Now other groups
are preparing to announce that they have succeeded in
creating other basic
computing components at this ultramicroscopic scale, known
as molecular
electronics.
Researchers at Yale and Rice Universities, for
example, plan to report in the journal Science in
a few weeks that they have taken an important
step past the Hewlett-UCLA work. In the July
demonstration, the molecular gate could be
made to move into open or shut positions, but
could not be switched back again. But the
Yale-Rice team says it has created
molecular-scale switches that can be repeatedly
opened and shut -- a necessary step in
representing zeros and ones, the basic binary
signals used in the circuitry of digital transistors.
And now Hewlett-Packard scientists say they
have recently taken an important step toward
creating rows of conductive wires that are less
than a dozen atoms across -- a crucial part of
hooking together the molecule-sized switches
that could one day result in computers vastly
faster than today's.
The rapid sequence of breakthroughs is giving the
researchers a new sense
of confidence. "We're on the scent, and we know the fox is
out there," said
Stan Williams, a Hewlett-Packard physicist who is a pioneer
in molecular
electronics.
According to the buzz in this research community, meanwhile,
other
laboratories are making progress on a number of fronts,
working under
top-secret conditions. One of these labs is said to be
working on a molecular
device capable of holding random-access memory, or RAM.
If molecular memory devices could be constructed, they might
offer vast
storage capabilities for just pennies in cost. One near-term
application might
be to permanently store an entire DVD-quality movie in a
space much smaller
than a conventional semiconductor chip.
As an applied science, molecular electronics would begin at
a minute scale far
beyond the theoretical boundaries of the conventional
technology of silicon
transistors.
Today's silicon-based
microelectronic devices have a
minimum size between electrical
components of 0.18 micron
(about one-thousandth the
thickness of a human hair) and
could potentially go as small as
0.10 micron. That would be
100-billionths of a meter, or 100
nanometers.
But in molecular electronics, the
smallest components may be
able to shrink to one-hundredth
that size -- a single nanometer.
The difference could mean chips exponentially more powerful
than anything
of a comparable size today or computing devices unimaginably
tiny by
contemporary standards.
The recent rapid pace of advance has led to a palpable sense
of mission
among a small group of physicists, chemists and computer
designers, who
until recently were viewed as impractical dreamers by much
of the computer
industry.
"In two to five years, you will begin to see functioning
circuits which are of
recognizable utility," said John Ellenbogen, a molecular
electronics researcher
at Mitre Corp., a research center for the military and
private industry.
Such optimism leads a number of researchers to believe that
rapidly
cascading advances in molecular-scale science may soon
constitute what
economists refer to as a disruptive technology -- one that
changes basic
industrial assumptions, just as the transistor did in
replacing the vacuum tube
during the 1950s, and as integrated circuits overtook
individual transistors
during the 1960s. Some molecular electronics researchers
envision an entirely
new industry, perhaps within the next decade.
The consequences of such a revolution would be immense and
possibly
destabilizing for the world's semiconductor industry.
Although the chip
industry now believes that it sees a path at least until
2014 for making
ever-smaller solid-state silicon devices, the cost of the
manufacturing
systems needed to make the chips is enormous -- and
continuing to mount
with each new chip generation.
Today's semiconductor chips are made in multibillion-dollar
fabrication plants
-- or "fabs" -- that use light waves to etch successive
layers of circuitry on a
silicon substrate. It is an expensive process, in part
because of the high cost
of creating and maintaining the "clean rooms" required for
avoiding
contamination by dust. But researchers in molecular
electronics are optimistic
that they will be able to use much less finicky methods by
creating chemical
reactions that "self-assemble" vast numbers of
molecular-scale circuits at
infinitesimal cost.
"This should scare the pants off anyone working in silicon,"
said Mark Reed,
a Yale University chemist, who is co-author of the
forthcoming Science
article and co-leader of a related memory project to be
announced Dec. 6 at
the International Electron Device Meeting in Washington. "It
will be dirt
cheap and it will create a discontinuity."
A colleague in the field agrees. "If you can make computers
as easily as
photographic film, then a lot of companies are going to be
wondering what
they're doing with their $15 billion fabs," said James
Heath, a UCLA chemist
who is part of the Pentagon-financed Hewlett-Packard-UCLA
research team
that demonstrated molecular logic gates last summer.
The vision of a new industry
has captured government and
corporate attention. The Clinton
administration is now
considering the possibility of a
National Nanotechnology
Initiative as early as next
January to set up financing and
help organize diverse research
activities in nanotechnology -- a
range of manufacturing
technologies that begin at the
scale of individual molecules.
Moreover, a number of
computer and semiconductor
companies, led by Sun
Microsystems and Motorola,
have been quietly meeting with scientists to discuss the
formation of an
industry consortium to seek commercial applications for
molecular
electronics.
And yet, researchers acknowledge that so far they have taken
only the first
baby steps toward the larger challenge of building
molecular-scale
computers. No one, for example, has figured out how to
interconnect billions
and billions of molecular switches with wires 11 atoms in
diameter.
"It feels like we're a year before the invention of the
transistor and we're
asking: 'What does solid state look like?"' said Paul Saffo,
a researcher at the
Institute for the Future who has tracked the development of
new
technologies.
And there is a general agreement that if such systems are to
be assembled
into workable computers, it will require radically new
architectures alien to
today's semiconductor-based computers.
At Hewlett-Packard, at the Massachusetts Institute of
Technology's
Laboratory of Computer Science and at Mitre, computer
scientists are
beginning to explore computer architectures that are far
more fault-tolerant
than today's microelectronic computers and whose structures
resemble
biological systems.
Manufacturing might involve assembling trillions of circuits
and then
identifying and mapping out the bad ones -- much as faulty
sectors are
declared off limits in today's disk drives.
"We will try to program with what we've got," said James
Tour, a molecular
scientist at Rice University. "It's a very biological
approach. Everyone's brain
is the same, but the pathways are all unique."
Much of the research financing in this field now comes from
the Pentagon's
Defense Advanced Research Projects Agency. "We've built this
entire
program on the idea of thinking differently," William
Warren, a program
manager at the agency, said. "We don't want to be standing
on the shoulders
of silicon."
That is why the recent first steps toward eventual
self-sufficiency have
created such excitement in the molecular computing research
community, a
group that for years had a consistent vision but no
empirical results.
"All of us were constantly on the defensive," Ellenbogen,
the Mitre
researcher, recalled. "Although we believed in some rational
way this was the
way to go, among ourselves we were continually forced to
reassure
ourselves that we weren't crazy."
By JOHN MARKOFF
Copyright 1999 The New York Times Company
http://www.nytimes.com/library/tech/99/11/biztech/articles/01nano.html
janet paterson
52 now / 41 dx / 37 onset
613 256 8340 po box 171 almonte ontario canada K0A 1A0
a new voice: <http://www.geocities.com/SoHo/Village/6263/>
<[log in to unmask]>
