Under the leadership of cofounder Scott McNealy, Sun Microsystems, Inc.,
has, according to Fortune, become one of America's fastest-growing
companies. The 39-year-old chief executive is heading "the most
efficient company in the industry," says BusinessWeek, which also
credits McNealy with having "proved to his doubters that he can run a
multimillion-dollar company by making Sun Microsystems the model for the
entire industry, including IBM."
McNealy has become an industry spokesman for open computer systems and a
chief proponent of client-server, peer computing. In addition to his
duties as CEO, McNealy serves as chairman of the board of Sun, the
parent company of all Sun subsidiaries and operating companies. The
$4.3-billion corporation, which ranks 139 of the Fortune 500 companies,
is the world's leading provider of power UNIX workstations, servers, and
related software and hardware technologies based on open, distributed,
network computing.
Educom Review: Let's start by talking about some marketplace issues and
then go on to some more general questions. With pressure from
Windows/NT, the UNIX industry has recently espoused standards such as
COSE [Committee for Open Systems Environment], but many people now view
the desktop as a commodity market. Can Sun be successful in such a
market?
McNEALY: First, COSE is a response to consumers, not a reaction to
anything happening in a little town in the state of Washington.
And what consumers are demanding is open, scalable systems. Sun has been
successful for 13 years not only because of its technological edge, but
because it has driven open computing. The reason SBus, NFS [Network File
System], and other technologies from Sun have become standard is that
they are good technologies, and they are open.
Open interfaces, or what we refer to as barrier-free interfaces, are
specifications that anyone has access to. They are not proprietary. Now,
the implementations based on those open interfaces are and should be
proprietary. No one holds a copyright to the English language, yet an
author can arrange letters and words in a particular order that you
perceive as an added value. So you're willing to pay for that
implementation. That's the difference.
Any parent who has purchased a Nintendo game for the kids knows what it
feels like to get locked into proprietary technology, because after
plunking down a couple of hundred bucks for the machine, they can buy
the games only from Nintendo. There's no competition, no innovation, and
no choice for the customer.
Business managers understand this. They understand that buying
proprietary technology is costly. And that is a major advantage for Sun.
It has propelled us from zero to $5 billion in revenues in 13 years. I
can't think of better evidence than that open standards make good
business sense.
That's why COSE has embraced a philosophy that Sun pioneered. The
consortium, representing a fair share of the UNIX world, sees how open
systems benefit the consumers as well as the suppliers of the
technology.
As for the desktop wars, Sun has made it very clear we do not intend to
compete by making low-end boxes that run a spreadsheet or word
processing program--one at a time--with networking as an afterthought.
We compete by being the company that you run your business on. That
means multitasking, networked, scalable, open SparcStations and
SparcServers, running Solaris, the most robust operating platform in the
industry. And then we'll run your spreadsheet and word processing
programs, too. You can run more than 8,000 UNIX applications or your
favorite Windows or Mac applications. With machines starting at $2,995
each, we are more than competitive, especially when you take into
account total cost of ownership in maintaining those systems.
ER: Your collaboration with NeXT and recent announcement with Fujitsu
seem to suggest that Sun needs to collaborate to survive.
McNEALY: What is clear, what is driving the consolidation in the
industry is that it costs a lot of dollars to play in the computer
business today. Back in the old days, when we did NFS, we did it with
about three engineers and half a marketer. You can't do that anymore.
When we did Sparc, the first implementation cost about $20 million to do
the first ASIC [Application Specific Integrated Circuits] version and
come out with a first computer. And it now outships all the other RISC
[Reduced-Instruction-Set Computer] chips combined.
How much money has Digital Equipment Corp. spent on its Alpha chip? How
much have IBM and Apple spent on the PowerPC? How much has Microsoft
spent on NT? So what's driving consolidation is the cost of delivering
in high enough volume to reap a return on the investment.
The dollars spent to create new interfaces and drive new interfaces into
the marketplace in terms of engineering, evangelism, investment,
marketing, brand recognition, vaporware--all the other things you've got
to do these days to make a technology successful--make it just so
expensive that you need incredible unit volume.
Our agreement with NeXT provides Sun with a key element in our Solaris
Distributed Object Environment effort. It will speed our time to market
with object-oriented programming. But it also goes beyond a licensing
agreement. Working with NeXT, we have established the API [Application
Programming Interface] for OpenStep as an industry standard.
This is a key win for open computing in a market that will drive a key
part of the computer industry in the '90s.
ER: What about a public windows interface? What's the status of Sun's
hopes to create an independent standard for PWI? Is the grassroots
approach paying off?
McNEALY: It's not a question of a hope on the part of Sun or anyone
else. The PWI initiative was a response from strong consumer demand.
People were just plain tired of one company maintaining such tight
control. Microsoft didn't seem interested in opening up its APIs, so we
did it for them.
ER: There are a lot of anecdotes on the net from sites that say they had
difficult transitions to Solaris 2. Has Sun given credence to those
stories by announcing continued support of SunOS on many newly announced
systems?
McNEALY: Over 75 percent of all applications for the Internet--Gopher,
WAIS, Mosaic--were developed on Suns. It clearly means, "The network is
the Sun computer." To answer your question, the fact is that Solaris 2
in today's version [2.3] is the best industrial implementation of UNIX
delivered in high volume. Tune your Sun workstation to the net again and
listen. The robustness of Solaris 2.3 is at the same level as the latest
release of SunOS 4.1.3.
Solaris 2 is also expanding in technology with new features, such as a
multithreaded kernel that allows users to take advantage of Sun's
multiprocessor hardware and such administration tools that now include a
serial port manager to make it easy to administer modems and terminals
across the network. And this is a reality.
At the same time, our customer base is in the process of changing over.
It is a million-unit, installed base, and when we initially planned our
changeover, we were overly optimistic as to how fast our customers could
change. We also experienced the migration to Solaris 2 ourselves, and 98
percent of our internal installed base is running Solaris now [24,000
workstations and 2,000 servers].
ER: Sun now supports Solaris 2 on Intel-based platforms. Is Sun becoming
more of a software company? How steadfast is its support for the Sparc
microprocessor?
McNEALY: Take a look at our recent announcement for Sparc, in which Sun
and Fujitsu have mapped out a very long term commitment to Sparc's
research and development. Our investments, when combined with our
fabrication partners, exceed $500 million a year.
ER: What effect will the Pentium and PowerPC chips have on the desktop
workstation market? Can Sparc keep up with performance gains from these
microprocessors?
McNEALY: The unified road map between Sun and Fujitsu will provide
enhanced 32-bit and advanced 64-bit processors that will maintain a two-
to-four-times performance advantage over anything Intel puts out there.
As for the PowerPC, it remains to be seen. We can't discount the
multimillion-dollar advertising budget of IBM/Apple, but let's look at
the chip on its technical merit: it has no installed base compared with
the 1 million units for Sparc; the architecture is proprietary to IBM
whereas Sparc is open; the chip has virtually no applications, compared
with 8,000 for Sparc; and it has neither the scalability of notebook to
supercomputer nor the multiprocessing capabilities of Sparc.
ER: Sun's role in education has been focused primarily on power users,
such as engineers and scientists engaged in research activities. Do you
see a wider role for Sun's participation in the education market?
McNEALY: The answer is definitely yes. We are strongly investing our
collaborative efforts in instruction technology by covering a total of
four segments: Computer Sciences, Instructional Technology and
Multimedia, Digital Information and Campus Administration, and
Biomedical and Health Care.
For each of these segments, our objective is to cover both research and
instruction. Our background as a key provider in the engineering-
scientific market has made research a natural environment for us. And we
are very competitive in the price-performance arena, which makes us very
attractive to teachers. Today you can equip an entire classroom with
SparcClassic workstations for less than the price of a PC- or Mac-based
solution.
And with Sun you have the advantage of integrated standards and network
connectivity. Students from grade school on up are connecting to the
Internet via Mosaic [another tool developed on a Sun].
ER: How have you changed in the past 10 years?
McNEALY: My time horizon is longer when making decisions. Ten years ago
I was involved in decisions that affected Sun in a matter of weeks; now
I deal with questions that will affect us for years. I have gotten
engaged, my weight has crept up about a pound per year, and I am on the
road more. I am smarter, but slower at hockey.
My focus is more on the customer and less on the competition these days,
but I will always believe computer interfaces should be as open as
English. And until all children are safe from proprietary interfaces, my
job will be fun, challenging, and rewarding.
ER: As an industry leader, what is your biggest frustration? your
biggest disappointment? your biggest hope?
McNEALY: As a representative of Sun Microsystems, my job is to act on
behalf of the shareholders and employees. That can be frustrating
because as a public figure I have to keep my personal opinions to
myself--okay, I try to keep them to myself.
My primary hope is that the world does not forget the value of smaller
government, free market economies, and personal responsibility. The
evidence shows the tragic results of big government in which citizens
expect everything to be done for them. That's why communism failed. But
making capitalism work requires citizens who take responsibility for
themselves.