DON'T BE LED astray by Carly Fiorina's mug on the
cover of BusinessWeek. Hewlett-Packard
(HWP) was slow to catch on to the computing
revolution, largely because since the company's
founding in 1939 in a garage in Palo Alto, Calif.,
it has made a profitable business off of some of the
most important -- and the most obscure -- technology
of the second half of the 20th century. As HP
proper prepares to push its e-commerce strategy and
compete with Sun Microsystems (SUNW) as a
supplier of Internet computing systems, Agilent, the
crazy uncle in the attic, stands to become a
first-rate communications technology company, up
there with the likes of Lucent (LU), Conexant
(CNXT) and other greats.
What everyone's excited about, of course, is that
Fiorina will turn HP's computer business into a
stock like Sun Microsystems -- suddenly an Internet
player, in other words. HP could do it. Its Unix
systems are among the best in its class -- better
than Compaq's -- and it has worked closely with
Intel on the development of the latter's
next-generation Merced architecture. With a growing suite of
storage-networking products, HP's servers could
compete with Sun in the growth of Internet
infrastructure inside of ISPs and the new class of
application service providers.
But everyone's forgotten about "Test." This week, HP
spun off its venerable test and measurement
division with a splashy naming ceremony, dressing up
the group, which it hopes to sell to the public
this fall, as something called Agilent. Despite the
fanfare, Agilent has been regarded as the ball and
chain around HP. Its gross margins, which
NationsBanc Montgomery analyst Kurtis King estimates
are in the 43% range, have supposedly hamstrung the
computer division from slashing its 30% gross
margin to the bone in the cutthroat PC market. The
profitable Test is screwing up HP's brilliant opportunity to lose money,
in other words.
So what the heck is this test stuff that's been
clogging up HP? Only some of the most important parts of the
communications revolution. Yes,
parts of Agilent are slow growing. King estimates
that Agilent represents a 9% growth rate between 1994 and 1998, vs. 19%
for HP proper. And
yes, some of the stuff is really obscure. The test
products themselves, where HP truly excels, take the company back to its
roots in the 1930s,
when Dave Packard and Bill Hewlett cut by hand their
own instrument faceplates out of sheets of metal.
Back then, Hewlett and Packard were both experienced
students of radio technology, and their first product was something
called an audio
oscillator, a device used to generate a precise
frequency signal for various communications applications. (One was used
by Disney to engineer
the sound for Fantasia.) Over the years HP pioneered
devices such as oscilloscopes. Today, HP's oscilloscopes are used for
high-tech
communications operations such as testing signal
integrity for fiber-optic networks. HP has tests for dozens of important
areas of networking. Its
machines can test the frequency range of new
wireless networks. And it offers products to test Rambus (RMBS) RDRAM,
which some still
think will be the global dynamic memory standard in
the computer business for the next several years.
All that is important, and it's very, very
high-margin, which is nice. But what's even better, and what gets left
out when people say "test and
measurement," is the part of HP devoted to various
electronic and optical components of all shapes and sizes. These
products, which figure in
just about every aspect of the communications
revolution, could turn HP into the equivalent of JDS Uniphase (UNPH),
the fiber-optics giant
whose stock has climbed from 34 7/16 at the
beginning of the year to 90 3/8 on Friday, based on an amazing store of
optical-networking
ingredients. Basically, HP has many of the most
important parts needed to build today's broadband, high-speed networks.
That could make it a
stunning new pure play in communications come the
fall.
The Semiconductor Products Group of Agilent, which
is separate from the test division, has all the goodies. On the
corporate networking side,
the company produces 1,300 nanometer lasers that can
be used for high-speed local-area networking of the Gigabit Ethernet
variety. The
company also has a suite of computer chips for the
storage technology called Fibre Channel. Storage networking is hot, of
course, because the
computer division of HP, along with IBM (IBM), Sun,
etc., have started to compete with the reigning king of storage
networking, EMC (EMC).
Fibre Channel is simply the next act, and newly
public companies such as Brocade Communications (BRCD) could use
Agilent's chips to
make their equipment.
HP also has wireless amplifier chips for the CDMA
cell-phone market, and chips for higher-speed wireless broadband
networking in the so-called
millimeter wave range, at speeds above 30 gigahertz.
Millimeter wave is a candidate for phone and Internet access systems in
urban markets
where young, competitive carriers want to bypass the
wired infrastructure. The company has also developed high-speed
transmitters for
fiber-optic networks at speeds of 2.5 gigabits per
second, as well as various types of so-called pump lasers for optical
amplifiers that could be
used in various types of all-optical networking.
And then there is HP's longstanding work in
light-emitting diodes. The company spun out the highest-performing LED
technology into a 50-50
joint venture with Phillips of the Netherlands,
called LumiLeds, while Agilent is retaining the LEDs used as indicator
lights for consumer
appliances and industrial equipment. The company is
also an intellectual-property holder in the vertical-cavity
surface-emitting laser (VCSEL)
field, which is starting to heat up, thanks to the
need for high-speed semiconductor lasers in computer networking.
In other words, Agilent has a brilliant base of
products on which to move forward. Where it goes from here depends a lot
on how CEO Ned
Barnholt decides to position the company. Its
medical products, such as its sonogram meters, are apparently
world-class, and I doubt they'll
drag down the rest of the company. But if I were
Barnholt, I wouldn't stop talking about how this is really a
communications pure play, with a
finger in every pie.
Supposedly, come November the split will be formal
and everyone at HP and Agilent will be carrying the right business
cards. Agilent has not
yet registered for a public offering, but that could
happen by September.
I'm still stumbling over the name. Unlike the
decade's most miraculous naming effort, Lucent, Agilent smells too much
of high-paid consulting.
Plus, it rhymes with a bodily function. But the more
vexing question is what the right marketing tag line will be for a
test-and-measurement outfit
with ambitions to be an infrastructure provider for
the Internet. What you don't want is something like "We're the test in
Dot Com." Nor will "We
make the stuff that makes plumbing possible" carry
much weight with the average investor. No, the management team had
better put test and
measurement behind Agilent and start talking about
what a brilliant communications company this can be. |
