EVENTS, ADVANCES, AND NEWS
■ BY JEFF ECKERT
■ Uzi Vishkin’s prototype desktop
PHOTO COUR TES Y OF CLARK SCHOOL OF ENGINEERING,
UNIVERSI TY OF MARYLAND.
For many years, the computer
industry relied primarily on chip
fabrication improvements to produce
higher and higher clock rates, thereby
keeping Moore’s Law alive. But
physical limitations have begun to kick
in, and it looks like the best way
to achieve greater computing power
is through parallel processing. This
is exemplified by the newer dual-and quad-core processors that are
appearing in PCs, but it appears that
Professor Uzi Vishkin, at the University
of Maryland ( www.umd.edu), has
leapfrogged ahead of the game.
Vishkin and colleagues at UM’s A.
James Clark School of Engineering
have developed a prototype machine
that uses 64 parallel processors, thus
promising to provide speeds 100 times
higher than current desktops. Vishkin
has actually been working on the concept since 1979, when he began developing a theory of parallel algorithms
that relied on a mathematical model of
a parallel computer (at the time, no
viable parallel hardware existed).
By 1997, advances in technology
enabled him to begin building a
prototype desktop device to test his
theory, and it was completed in
10 November 2007
December 2006. He kept it under
wraps until recently but finally, based
on positive reactions from graduate
students and others, decided it was
time to let the world in on it.
For example, according to
Charles E. Leiserson, professor of computer science and engineering at MIT,
“The single-chip supercomputer prototype built by Professor Uzi Vishkin’s
group uses rich algorithmic theory to
address the practical problem of
building an easy-to-program multicore
computer. Vishkin’s chip unites the
theory of yesterday with the reality of
today.” Vishkin believes that his
approach (on which he holds several
patents) will eventually lead to a revitalization of the computer industry.
FISH AND LEDS SMELL
IN THREE DAYS?
PHOTO COURTESY OF US GEOLOGICAL SURVEY.
■ This grumpy-looking Chinook salmon
perhaps suspects his fate.
As reported here from time to time,
there is a continuing quest for
brighter LEDs. It now appears that, in
collaboration with the US Air Force
Research Laboratory ( www.wpafb.af.
mil/AFRL/), Professor Andrew Steckl
of the University of Cincinnati
( www.uc.edu) has come up with an
interesting (and somewhat distasteful)
As we know, an LED emits a photon whenever negative and positive
charges recombine in the P region of
the semiconductor. But some of the
electrons passing through the device
fail to find a positive mate and are
therefore wasted. A challenge has
been to find a better way to manipulate the electron’s mobility so as to
trap them longer and give them more
time to latch onto a positive charge.
In his search for cheap, plentiful
materials that can be used for such
purposes, Steckl has discovered
salmon DNA. To be specific, salmon
sperm. He noted, “Salmon sperm is
considered a waste product of the
fishing industry. It’s thrown away by
the ton. It’s natural, renewable, and
The result is devices dubbed
BioLEDs, which incorporate DNA thin
films as electron blocking layers.
Standard devices use inorganic materials such as silicon but, according to
Steckl, DNA’s unique properties allow
two orders of magnitude improvements in efficiency and brightness.
Research continues, with the long-term goal of making “green” devices
that use only natural, renewable, and
The professor’s work seems to be
generating considerable interest. “I’m
receiving salmon sperm from
researchers around the world wanting
to see if their sperm is good enough,”
he observed. While Steckl is currently
focusing on salmon, he thinks that
other animal or plant sources might be
equally useful, so keep your dogs and
cats inside at night.
AI LAB BATTLES
Many recent news features (and
legal controversies) have
focused on passive monitoring of
communications between suspected