1. Chip sales up 31% in 2000; Toshiba passes NEC as No. 2 supplier
  2. Japan's chip makers remain cautiously optimistic about New Year
  3. DNA-based sensor chip detects metals in real-time
  4. North American Semiconductor industry reports 1.12 book-to-bill ratio for November

Chip sales up 31% in 2000; Toshiba passes NEC as No. 2 supplier
(01/02/01 11:48 a.m. EST)

SAN JOSE -- Semiconductor sales increased 31% to $222.1 billion in 2000 from $169.1 billion in 1999 despite a slight slowdown in revenue growth during the final months of the year, according to preliminary estimates released today by Dataquest Inc.

Dataquest's initial 2000 market report also shows Toshiba Corp. overtaking NEC Corp. as the world's No. 2 chip supplier in 2000. For most of the 1990s, NEC had been the world's second-ranked semiconductor company behind Intel Corp., but Toshiba's strong 47.2% growth in chip sales pushed it just slightly ahead of its Japanese rival, said Dataquest.

Toshiba's chip sales reached $11.2 billion in 2000 from $7.6 billion, estimated the San Jose-based research firm. Toshiba was ranked No. 3 in 1999. NEC's semiconductor sales grew 20.3% to $11.1 billion in 2000 vs. $9.2 billion in 1999, Dataquest said (see table below).

While semiconductor sales grew at a strong rate in 2000, Dataquest noted that the chip industry is "having trouble" as it moves into a new year. Slowing shipments of key applications--such as personal computers and cellular phones--coupled with inventory adjustments delayed new orders in the second half of 2000. Some analysts have expressed concern about too much chip capacity coming on stream as demand cools. Most market research firms have chopped their 2001 forecasts for growth to about half of last year's increase, with some analysts calling for just over 10% increase in many key IC segments.

"We are now at the end of the second year of an up cycle and there is despondency regarding the future, as the fourth quarter of 2000 was weak relative to the preceding quarter," said analyst Joe D'Elia, vice president and director of Dataquest's European semiconductor research. "Historically, the industry has gone through inventory corrections during the positive portion of the industry cycle, and we see no reason to believe that this current weakening is anything else other than an inventory correction."

Intel continued distance itself from other chip suppliers in 2000. The Santa Clara, Calif.-based company had semiconductor revenues of $29.8 billion in 2000, an increase of 11% from $26.8 billion, according to Dataquest. Intel's sales in 2000 were $18.6 billion higher than Toshiba's revenues in 2000. In 1999, Intel's sales were $17.6 billion higher than NEC.

In the Top 10 ranking, Europe's STMicroelectronics moved up to No. 7 from No. 8 with a 56.5% increase in revenues to $7.9 billion in 2000 compared to $5.1 billion in 1999, Dataquest said. Hyundai Electronics Industries Co. Ltd. of South Korea jumped into the Top 10 at No. 9 from No. 11 in Dataquest's preliminary ranking. Hyundai's chip revenues grew 42.6% to $6.9 billion from $4.8 billion, Dataquest estimated. Infineon Technologies AG slipped to 10th place from eighth with sales of $6.7 billion vs. $5.2 billion in 1999.

How top chip makers ranked in 2000

2000 1999 Supplier 2000 sales % increase 1999 sales 2000 share
1 1 Intel $29.750 billion 11.0% $26.806 billion 13.4%
2 3 Toshiba $11.214 billion 47.2% $7.618 billion 5.0%
3 2 NEC $11.081 billion 20.3% $9.210 billion 5.0%
4 4 Samsung $10.800 billion 51.6% $7.125 billion 4.9%
5 5 TI $9.100 billion 27.8% $7.120 billion 4.1%
6 6 Motorola $8,000 billion 25.1% $6.394 billion 3.6%
7 9 STMicro $7.948 billion 56.5% $5.077 billion 3.6%
8 7 Hitachi $7.282 billion 31.1% $5.554 billion 3.3%
9 11 Hyundai $6.887 billion 42.6% $4.830 billion 3.1%
10 8 Infineon $6.715 billion 28.6% $5.223 billion 3.0%

For more technology news, visit http://www.siliconstrategies.com

Japan's chip makers remain cautiously optimistic about New Year
(01/02/01 14:52 p.m. EST)

TOKYO -- Electronics production in Japan will hit a record high this year and will continue to grow by 7.1% next year, according to projections by the Japan Electronics and Information Technology Industries Association (Jeita).

Jeita estimated that total electronics production in Japan, including consumer electronics, industrial electronics and general-purpose devices and components will reach about $231.4 billion this year, up 10.1% from last year. This will slightly surpass the past peak of about $231 billion in 1997.

The growth this year is mainly due to the large increase in semiconductors and display devices, which account for 45% of the total electronics production and which have grown 16.8% from 1999.

Jeita predicts that in 2001 the growth rate will be less than this year, but production will still increase by 7.1% to $248 billion. Fastest growth areas will include information technology (IT) related products and digital consumer electronics. Production of semiconductors and LCD panels is expected to show continuous two-digit growth.

In November and December 2000 manufacturers have reported some negative trends in orders, but Hiroshi Tsukamoto, president of Jeita, remarked, "in a sense, the growth until autumn this year was too much." He said that currently the industry is in an adjustment phase and is headed toward a brisk turnaround again next spring, "supported by a steady demand especially from the IT arena and cellular phones." In another survey, Jeita projected that the world cellular phone market will grow to 453 million units in 2001, up 19% from this year's 380 million units.

Jeita is the combined organization of the former Electronic Industries Association of Japan and the Japan Electronic Industry Development Association.


For more technology news, visit Electronic Buyers' News

DNA-based sensor chip detects metals in real-time (12/29/00, 7:35 a.m. EST)

CHAMPAIGN, Ill. — An inexpensive, real-time sensor technology harnesses living DNA to detect dangerous metals such as lead, mercury and cadmium.

Developed by researchers at the University of Illinois, the DNA sensors immediately react to the presence of specific metals by emitting light into an inexpensive fiber-optic lens. Traditional methods require lengthy batch testing or expensive instrumentation. Genetic algorithms were used to discover the specific required DNA strands required to detect specific metals from within a population of trillions of random DNA sequences.

Engineers could benefit from this method by designing their own DNA strands that would test in real-time for specific metallic substances.

"We have created a new class of simple and environmentally safe sensors — the world's first example of a catalytic DNA-based biosensor with highly sensitive fluorescence detection for metals," said professor Yi Lu. Lu was assisted by graduate student Jing Li, co-author of their recent paper in the Journal of the American Chemical Society. The process patent is pending.

Lu's technology harnesses the human body's ability to detect specific kinds of molecules and orchestrate responses in real-time. This molecular recognition capability involves a "lock and key" mechanism in which custom-tailored receptor "pockets" will only accept specific molecules, and respond immediately when they do encounter the correct molecules. While the human body grows these specific receptor sites according to the blueprints contained in DNA, Lu decided to go straight to the source. "Our sensor's technology is unique, because the active element consists of small pieces of real DNA, the basic building block of all our genes," said Lu.

Lead and other dangerous environmental contaminants, such as industrial mercury and cadmium, can be detected today only after lengthy batch testing of samples for specific elements. Thus there is a need for a quick and inexpensive method for on-site, real-time testing for hazardous substances.

That capability has become a high priority at the National Institutes of Health, which provided the funding for Lu's experiments. The NIH has specifically targeted health applications for the technology, including environmental monitoring, clinical toxicology, wastewater treatment and industrial process monitoring.

Lu's innovation is based on a 1994 pharmaceutical discovery that DNA was not just a genetic information repository, but that DNA could also act in a manner similar to living enzymes that catalyze a specific chemical reaction right at the site where it is needed. As a result of this discovery, many promising new pharmaceutical agents have been demonstrated in which metal ions are essential to activate the catalytic function. These enzymes, called catalytic DNA, constitute a new class of metalloenzymes derived from metallo-nucleic acids. Lu's innovation was seeing how to turn this pharmaceutical discovery into a sensor technology.

Lu engineered a way to attach a fluorophore to one end of the DNA strand and a fluorescence quencher on the other end. In steady state when exposed to 560-nm light for excitation of the flourophore, the quencher's proximity keeps the fluorophore from glowing. But when the desired metal is present — lead in his demonstration experiment — it cleaved the quenching end, resulting in an easily detectable 400 percent increase in fluorescence. "DNA is stable, cost-effective and easily adaptable to optical-fiber and chip technologies," said Lu.

To turn his discovery into a working sensor technology, something the researchers have not yet done, they would attach the fluorophore end to a chip substrate designed to have an optical fiber permanently attached. Then when the quenching end is cleaved, it leaves the 400 percent increased fluorophore glowing directly into the fiber optic, indicating the presence of the hazardous metal. These chips could also be "reset" by washing off the sample and chemically reattaching new fluorescence quenchers for the same or a different metal to be detected.

To turn his theoretical biosensor into a working technology, Lu had to turn to genetic algorithms. While the 1994 discovery of catalytic DNA demonstrated its principle, now six years later, the process by which the base-four DNA codes (using A, C, G and T "bits") respond to specific molecular shapes is still a mystery. Theoreticians continue to speculate on the specific mechanisms at work, but an analytic understanding that would allow designers to specify particular A, C, G and T sequences as "locks" for specific metal "key" molecules is not yet possible. Rather than wait for the theoreticians to hash it out, Lu responded in typical engineer fashion, by appealing to genetic algorithm searching techniques to merely test millions of randomly generated DNA strands and select those that respond to the desired metal. Here is where Li took over doing the in-the-trenches work of generating trillions of different DNA stands in a test tube, then performing genetic algorithms to select the ones that happen to work for detecting lead.

Li implemented the genetic algorithm by first generating a massive population of 1 quadrillion random DNA strands — individual strands that can fold like proteins, rather than the familiar intertwined double-helix that serves as a blueprint for living things.

A natural selection process then filtered out those strands that could only fold around lead, pictured as a DNA "pocket" that can only fit a specific molecular shape. This smaller population was then subjected to random mutations which were multiplied using polymerase chain reactions to create another large population, and the process repeated.

In the end, Li was able to obtain DNA strands that could detect lead over a concentration range of three orders of magnitude. Since the fluorescence domain is decoupled from the metal-recognition domain, by virtue of the quenching effect, Lu has high hopes that Li will someday be able to run improved genetic algorithms that further advance not only the sensitivity of specific catalytic DNA strands, but also make sure that no untested metal ions are around that could accidentally trip the system

For more technology news, visit http://www.eetimes.com

North American Semiconductor industry reports 1.12 book-to-bill ratio for November (12/26/2000 )

The North American-based manufacturers of semiconductor equipment posted $2.74 billion in orders in November 2000 and a book-to-bill ratio of 1.12, according to the November 2000 Express Report published today by Semiconductor Equipment and Materials International (SEMI). A book-to-bill of 1.12 means that orders were 12% higher than shipments for the month.

The SEMI book-to-bill is a ratio of three-month moving average bookings to three-month moving average shipments for the North American semiconductor equipment industry. Shipments and bookings figures are in millions of U.S. dollars.

The three-month average of worldwide bookings in November 2000 was $2.74 billion. The bookings figure is 8% below the October 2000 level but is 62% above the $1.7 billion in orders posted in November 1999. The three-month average of worldwide shipments in November 2000 was $2.44 billion. The shipments figure is 5% below the October 2000 level but is 60% above the November 1999 shipments level of $1.53 billion.

The data contained was compiled by Arthur Andersen LLP, without audit, from data submitted directly by the participants. Additional information is available at www.semi.org.

For more technology news, visit http://www.siliconstrategies.com