• The development of ever more powerful yet less costly computer chips and lab chips, including some that
  partially self assemble;
• Petaflop computers; and
• Stoppable lightwaves that
  promise to multiply computing power by further orders of magnitude.

A projected 25 percent increase in worldwide chip sales this year (following
double-digit gains in 2003) is partly the result of the accelerating
digitization of consumer-electronic devices, but even more fundamentally a
result of innovation in semiconductor materials, design, and manufacturing
methods. The accelerating innovations, suggests an informative article by Olga
Kharif in BusinessWeek, account for the phenomenal current evolution in
TVs, music players, and cell phones, which are obsolescent practically on
launch. They are also responsible for acceleration in what Kharif may live to
regret calling “goofy” uses of chips, such as clothes with an MP3 player woven
into the fabric,* and carpets that sound an alarm if someone steps on them out
of hours.

Kharif writes at length about several innovations chipmakers are developing
to produce smaller chips with more functionality, including:

• X-ray lithography for etching nanoscale circuits (three to five years away),
• new materials (IBM is investigating 20 of them),
• integrating several chips into one (e.g., cell-phone chips that double as
  medical diagnostic devices),
• Implantable RFID (radio frequency identification) chips for
  counterfeit-proof ID (about a year away from mass production for one major
  manufacturer, but already on the market from Verichip),
• “Lab chips” (aka “biochips”) to diagnose diseases or test DNA (already on
  the market), and
• “Silicon dust” sensors to identify airborne pollutants (within a year) and
  eventually to carry drugs to kill cancer cells.

Chipmakers, says Kharif, “are spending on research and development as never
before.” Intel alone will spend $4.8 billion this year on R&D, up ten
percent from 2003. The consumer electronics manufacturers that consume the raw
chips expect consumers to replace their cell phones and other devices more often
(assuming that the cost of the chips continues to fall, as it has annually in
double digits so far), and see vast new markets opening as low prices make their
products sellable in second and third-world countries.

* See “Wearable Recharger” in the Devices
section.

Reference: Kharif, Olga (2004). “A
Whole New World of Chips.” BusinessWeek Online, January 21.

Biochip Roundup

BusinessWeek‘s Amy Tsao provides a snapshot of the current state of
the biochip industry. In a decade, the industry has gone from one company to
about a dozen or so and the chips have gone from holding a fraction of a genome
to six entire genomes on a single chip, and from analyzing the genome to
analyzing the proteome. Their price continues to fall while the revenue for
their manufacturers continues to rise, such is the demand. “Labs at drug
companies and universities can’t get enough of the latest versions,” she writes.

Two types of chip are vying for top spot: chips that are customizable to a
specific individual genome or drug target, and “catalog chips” holding the
genetic maps of organisms widely used in research, such as human, rat, and
mouse. One company, Agilent, apparently uses an ink-jet manufacturing process to
print custom chips on demand. A six-genome chip from Illumina, to be available
at midyear, will let customers run six of the same experiments on a single chip.
Industry leader Affymetrix continues to add to its large library of catalog
chips containing whole genomes of heavily researched organisms.

A predicted major growth area is in biochips for diagnostic tests. One under
development will detect in a patient variations in a gene known to affect the
body’s ability to process about 25 percent of drugs on the market, thus helping
to tailor a prescription for that particular patient, and marking a major
milestone on the road to totally “personalized” medicine. Such medicine,
suggests Tsao, is “almost certain to become [the norm] at some point, as
innovative companies come closer to unleashing the potential of biochips.”

Reference: Tsao, Amy (2004). “The
Healthy Promise of Biochips.” BusinessWeek Online, January 21.

Nanovessels for Lab Chips

Researchers from the Swiss Federal Institute of Technology and IBM Research
have found a way to make nanoscale, self-organizing, biocompatible vesicles
(containers) for labs-on-a-chip. They demonstrated that an array of millions of
100-nanometer containers can self-assemble in a matter of minutes. Biochips
using this technology could reach the market in three to five years.

Reference: Unknown (2003). “Biochip Holds
Millions of Vessels.” Technology Research News, December 29.

Next Gen Supercomputers

At the behest of the US Department of Defense, Los Alamos National Lab
scientists are developing “the optimal design” for the next generation of
supercomputers, which will operate at more than a million billion floating point
operations per second, or one petaflop.

Reference: Unknown (2004r). “Lab Starts Next
Generation Supercomputers.” Associated Press/Excite News, January 15.

See also the Defense Advanced
Research Project Agency website.

Quantum, Photonic AI

Tokyo University researchers have designed a lightwave neural network that
could process massive amounts of information such as needed for holograms. The
enormous capacity results from processing in parallel the vast frequency range
of the lightwaves. A prototype “could be done within a few years,” according to
the researchers.

Meanwhile, Harvard University researchers have trapped and held a light pulse
still for a few hundredths of a millisecond before letting it go on its way. If
this technique were integrated with the lightwave neural net it could be used to
make lightwaves interact with each other, potentially multiplying the processing
capacity and potentially leading to computing at the quantum level, by
manipulating single photons.

This is somewhat speculative and may be two decades from any sort of reality,
but when it happens, the intelligence of a machine with the information
processing power of these technologies will be somewhat more than astonishing.

Reference: Unknown (2004). “Colors
Expand Neural Net.” Technology Research News, January 2.

Reference: Unknown (2003). “Light Frozen in
Place.” Technology Research News, December 30.