SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Qualcomm Incorporated (QCOM) -- Ignore unavailable to you. Want to Upgrade?


To: Ruffian who wrote (24029)3/11/1999 8:11:00 AM
From: Brian Lempel  Read Replies (1) | Respond to of 152472
 
Breakthrough chip technology, ASICS:

March 11, 1999

WHAT'S NEXT
I.B.M.'s Chip Sandwich:
Packing in a Lot of Power
By MICHEL MARRIOTT
As computer manufacturers scurried recently to announce their plans for the new and much-ballyhooed Pentium III chip from Intel, another technology giant, I.B.M., announced what truly seems to be a breakthrough in chip design. Some electronics experts said I.B.M.'s new design could significantly hasten the evolution of computers and personal electronics devices, allowing them to become ever smarter and ever smaller.


I.B.M.

SLIMMING DOWN - The SA-27E will do the work of several chips.
--------------------------------------------------------------------------------

I.B.M.'s chip is officially designated the SA-27E but is more commonly called a system-on-a-chip. It is 18 millimeters on each side, about the size of a thumbnail. In essence, the new I.B.M. chip promises to do computing jobs that now require two or more chips, and do those tasks faster and at a lower wattage and cost. The results, I.B.M. officials say, will be a range of much more compact, more efficient and more versatile products, like personal computers, cellular telephones, household appliances and video games.

The first devices using these chips are expected to begin showing up next year, said Phillip Bergman, an I.B.M. spokesman.

"This is quite exciting," said Bijan Davari, vice president for development at the I.B.M. microelectronics division in East Fishkill, N.Y., and an I.B.M. Fellow. He said the chip's design was unique. It can pack up to eight times the processing power and four times the memory found in the PC's typically used today, which use many chips.

Every computer, whether a sprawling work station or the electronic innards of a washing machine or toy, must be able to store sets of instructions or programs (its memory capacity) and use those programs to accomplish tasks (its logic capacity). These functions are usually assigned to separate chips that communicate with one another.

For example, a Pentium-class computer uses a very powerful chip crammed with millions of transistors to process data. It also uses clusters of other chips to store programs, data and instructions.

But this arrangement, chip makers say, has always been inherently inefficient, making for bulkier computers and slowing a PC's operation by forcing the chips to create streams of data traffic -- with the inevitable bottlenecks -- in the connections among them.

For years, scientists have sought ways to combine logic and memory into one chip without compromising the performance of either.

"Having the processing power and data on separate chips was like having the materials you need to do your job in another office," Dr. Davari said. "This forces you to keep going next door to get what you need."

Dr. Davari said the new I.B.M. chip, combining high-performance logic and memory, was the first of its kind. It manages the combination of tasks mostly through an innovative approach that embeds memory components into a sliver of silicon, the common material of all microprocessors. That is done, Dr. Davari explained, by burrowing beneath the surface of the silicon and placing the chip's memory, as much as 32 megabytes, into tiny trenches. These "trench cells" are as much as 6 microns deep. (A human hair is about 100 microns wide.)

The chip's surface is then left flat and uncluttered for the placement of highly compact logic transistors. Using special copper circuitry as thin as 0.15 microns, the entire package remains very small and very efficient, I.B.M. officials said.

The new chip may cost less than current chips because of the consolidation of functions and the need for less complex motherboards and other related components.

"People have been working on this for a while," said Chan Suh, chief executive of Agency.com, a consulting company based in New York that specializes in interactive technologies and strategies. He said I.B.M. was the first to offer designers this innovation because of the company's use of special copper circuitry for the logic functions; the tiny circuitry allows more memory to be packed inside the chip.

Suh said the new technology, while not as groundbreaking as the shift from vacuum tubes to transistors in the 1950's, is still an important step to greater miniaturization. And making devices smaller is "the Holy Grail," he said.

When the new I.B.M. chip technology is considered along with recent industry steps to establish a common computer language for small electronic devices, Suh said, the dawn of the talking toaster appears to draw closer.

Rick Doherty, director of research for the Envisioneering Group, a technology testing and marketing company in Seaford, N.Y., said the I.B.M. chip would above all open a universe of new possibilities to designers, freeing them from the constraints of multichip designs. "This allows people to get ideas in product form and to market a lot faster," he said.

Doherty added that he could envision, for instance, a "Palm Pilot X" using the system-on-a-chip. It would possess the power of a Pentium III and have lots of memory storage space, yet run on a single small battery.

In about a year, an I.B.M. spokesman said, the company expects to see the chips first put to use in improving Internet routers and hubs. Placing the chips, which will be made by I.B.M., in personal computers is part of the second phase of the chips' deployment, said Bergman, of I.B.M. He noted that such a use was probably two to three years away. The chips, ready for custom designs next month, will be ASIC's, or application-specific integrated circuits. Such chips are commonly found in things like digital cameras and automobile air-bag systems.

The SA-27E will make ripples in the technology world, Suh said. For consumers, he said, "the price of computers will definitely come down, and people will be able to get more things around the house with computerization built in."



To: Ruffian who wrote (24029)3/11/1999 10:04:00 AM
From: engineer  Read Replies (1) | Respond to of 152472
 
cgi.pathfinder.com

Qualcomm vs. Ericsson

The Battle to Control Your Cell Phone

Julie Creswell

For much of the past decade, Dr. Irwin
Jacobs has been at war. It began in the
early 1990s when the
professor-turned-CEO of Qualcomm
discovered that a certain military
technology could be developed into an
effective standard to operate cell phones
for civilians. Instead of being praised for
his breakthrough, Jacobs battled skeptical
naysayers, most notably a
Swedish-based nemesis named Ericsson,
which was promoting its own competing
standard to control the guts and
interoperability of cell phones.

Their battle continues today, but the
stakes are much bigger. Qualcomm of San
Diego, a $3.3-billion-a-year company, and
Ericsson, with $22.7 billion in 1998
revenues, are now in the middle of an
international spectacle that's been
dubbed a "holy war." The victor will usher
in the next generation of international
wireless communications. And Qualcomm
desperately needs a win. Revenue grew
59% last year, and profits rose 18%, but
its volatile stock ended the year virtually
unchanged, down $15 from a high of $67.
(It's now trading around $72.) After a
round of layoffs, the company may now
be looking to unload its troubled
equipment-making business.

So Qualcomm is hoping for transcendence
in a new era of high-speed data
transmissions, where you'll access e-mail
and Internet sites and even conduct a
videoconference call on the Santa Monica
freeway with the same phone that tucks
inside the palm of your hand. For any of
this James Bond-like technology to work,
governments and global manufacturers
like Ericsson and Qualcomm need to
cooperate. Easier said than done.

Part of the reason your cell phone doesn't
work in, say, Iowa and Paris and Buenos
Aires is the fact that there was no
attempt to ensure compatibility of
networks in the nascent days of analog
cell phones or even more recently when
companies shifted to digital cellular
technology. The International
Telecommunications Union, an agency of
the United Nations, would like to avoid
that situation in this coming third
generation of cellular communications. So
it is considering assigning a family of
global standards for new wireless phones.
While the ITU has said it would like to
come to a decision by the end of March,
the process may be delayed by a legal
showdown between Ericsson and
Qualcomm.

There's a lot of money and market share
at risk. Demand is booming in the U.S.,
and wireless communications are
expanding rapidly throughout Latin
America and Asia. By one estimate, the
global market for handsets will grow from
$40 billion in 1999 to more than $58 billion
in 2004. No firm has as much at risk as
Qualcomm. Most of its peers, including
Motorola, Nokia, and Ericsson,
manufacture phones for more than one of
the existing versions of the three
standards that now carry voice over
wireless networks: CDMA, TDMA, and
GSM. Not Qualcomm. The bulk of its 1998
revenue came from licensing fees,
semiconductor chips, equipment, and
handsets (with partner Sony) that are
linked to Jacobs' CDMA.

Jacobs had to battle to gain any
acceptance of CDMA. Originally used for
top-secret government communications
because it was impervious to jamming,
CDMA, standing for "code-division multiple
access," tags packets of conversation
with a unique numeric code. Many
different CDMA conversations get carried
over the same frequency; then their
packets get sorted and delivered to the
appropriate party. Its wireless technology
is growing fast in the U.S. and has a large
presence in Korea and Latin America.

Still, TDMA, which stands for
"time-division multiple access," is the
most successful standard in the U.S.,
with backing from operators like AT&T
Wireless and SBC Communications. Then
there's the other, similar technology
called GSM (global system for mobile
communications), which has dominated
Europe since its introduction in 1991.
Ericsson and Nokia work with both
technologies.

As they worked to develop cellular
technology that would carry voice and
video, and allow cell users to access the
Web, the GSM and TDMA camps joined
together and created a plan for a new
standard called wideband CDMA. (Yes,
CDMA.) Meanwhile, Qualcomm juiced up
its existing technology and created its
own third-generation CDMA. The
technologies sound similar and in many
ways are similar, but arcane technical
differences make them incompatible. For
consumers this means one thing: cell
phones still won't work across countries
or continents.

While Ericsson has a strong presence in
the European cellular market, Qualcomm
isn't on the radar there. If the ITU
endorses two different CDMA standards,
Qualcomm stands little chance of
breaking into the market, as subscribers
to its network won't be able to use their
phones there. Not surprisingly, Qualcomm
is urging the international body to
approve a single standard that's
compatible with its existing CDMA
technology. "When you're introducing a
brand-new technology that has a
common base, CDMA, then don't have
two slightly different ones," pleads
Jacobs, 65. Qualcomm's backers have a
lot at stake too--Sprint PCS has spent
more than $5.5 billion for a U.S. network
that supports CDMA technology, and
retrofitting to adapt to a slightly different
ITU standard would be expensive.

Another group of carriers, including
Airtouch Communications in San
Francisco, which is being acquired by
Britain's Vodafone, has urged a
compromise. "There are billions of
handsets to be sold, and plenty of market
share to go around," says Craig Farrill,
vice president of strategic technology for
Airtouch. But Ericsson executives insist
the company has made improvements to
CDMA that would be lost if its standard
were made compatible with Qualcomm's
existing technology. Qualcomm, in turn,
says that it owns the intellectual
property rights to CDMA and won't license
them unless the competing standards are
converged.

Qualcomm hasn't completely stalled
because of this standards battle. It has
two new product offerings this year--a
thin cell phone and something called the
pdQ smartphone (a cell phone with a
built-in Palm Pilot organizer). But these
products have a lot of competition,
including the introduction of new Nokia
CDMA phones and a CDMA version of
Motorola's popular StarTAC phone. And
even if Qualcomm's new offerings are
hits, it'll still lag behind leading handset
manufacturer Nokia (see chart). As a
result, warns Michael Ching, a wireless
equipment analyst at Merrill Lynch, pricing
pressure could slow Qualcomm's revenue
growth to 16% this year.

But Qualcomm's long-term success
depends on its version of CDMA being
embraced outside the U.S. The best
battle plan may be the one least
palatable to Jacobs: a pact with the
enemy. In April, Qualcomm is scheduled
to meet Ericsson in a Texas court to
argue the patent lawsuit. But reports
surfaced in mid-February that the two
were close to agreement on a pact that
would give Ericsson access to
Qualcomm's CDMA technology, and
Qualcomm the right to use Ericsson's GSM
technology. Analysts say that such a
deal makes sense, but that both sides
may need to compromise about who owns
which parts of the technology.

Compromise, though, has never been
Jacobs' strong suit. After battling for
acceptance for years, striking a deal with
the enemy can't be easy. In late
February a Qualcomm spokeswoman
insisted to FORTUNE that the company's
position hasn't budged. That
stubbornness could be Qualcomm's
Achilles' heel. Key players like Japan's
NTT are siding with Ericsson. If it doesn't
change its tune, Qualcomm may end up
with only its lawyers to keep it company.