Michael, thanks for posting the forbes.com stories
.... for those who have not figured out how to click on a URL (James, twister, etc), here is all of the text about SUNW from forbes.com ....
Is Sun Ringing Up A (Telecom) Monopoly? By O. P. Malik
Scott McNealy, the 42-year-old chief executive of Palo Alto, Calif.-based Sun Microsystems, relishes his role as Silicon Valley's court
jester. Even though his swipes at Microsoft Corp. founder Bill Gates approach pathological levels, one has to admit that McNealy is a very
astute businessman.
In 1996, when McNealy paid $100 million to Silicon Graphics for the rights to Cray Research's SPARC-based supercomputer
business, many thought he was a fool to spend that kind of money on a technology that had limited use. After all, only
governments and NASA could really afford "big-iron" machines costing more than $10 million.
But McNealy wasn't fazed. His engineers stripped those monsters down and made them the cornerstone of a supercomputer class of
machines they called Enterprise 10,000 (also known as Starfire). So far, Sun has sold 1,000 of these machines at $1 million each. That's
revenue of $1 billion, or ten times McNealy's purchase price for Cray, in just three years.
A fluke? Hardly. In 1996 Sun bought a small U.K.-based company called Integrated Micro Products for $96 million. IMP at the time
made fault-tolerant computers that were used by telephone networks and stock exchanges. After Tandem Computers and Stratus
Computer, IMP was ranked third in the fault-tolerent server business at the time. It helped that IMP had already started using Sun
Microsystems' SPARC chips.
McNealy wanted to use the IMP technology to go after the $9 billion a year telecommunications server market. Sun now gets about $1.5
billion in annual sales from the telecommunications business.
"Telecommunication [sales] are the single largest component of Sun's total sales," says Anil Gadre, general manager of Sun's Solaris
platform. Telecom-related sales account for slightly more than 15% of Sun's $10 billion fiscal 1999 revenue. Telecom related sales are
projected to hit $2 billion in fiscal year 2000.
Sun already sells computers (Netra FTs) that provide greater than 99.999% reliability and availability.
The computing needs of telecommunication companies can be roughly divided into two main categories: OAMP ("operate, maintain,
provision and administer") systems, which actually run the telephone networks; and BS (business support) systems, including customer
databases and billing services. Sun Microsystems is an active and dominant participant in the OAMP space.
Sun's computers are becoming ubiquitous on the telephone networks. "Sun is as pure a telecom play as you can get," says Robert
Rosenberg, president of Insight Research, a Parsippany, N.J.-based telecom research firm.
Rosenberg isn't exaggerating.
To understand Rosenberg's point, consider what happens when you place a telephone call from New York to San Francisco. The line used
to dial from New York talks to an AT&T switch located somewhere in Manhattan, where a Sun machine routes the call to an AT&T central
office. From there, another Sun machine reroutes the call to San Francisco, where the whole process is reversed.
The call goes through in a matter of milliseconds. Multiply that by the millions of calls that are made every minute and one
simple fact becomes clear: You need an extraordinarily powerful computer that is fail-proof and very fast to run a reliable
telecom operation. In other words, you need Sun's Netra FT servers.
Regional Bell companies such as Bell Atlantic or BellSouth have a huge appetite for this type of computer. "Sun sells a ton of equipment
to old-fashioned phone companies," says Dan Taylor, principal at Boston-based telecommunication research firm Giotto Perspectives.
And Sun's success in the RBOC (regional Bell operating company) and old phone company markets has come at the expense of traditional
leader Tandem Computers, which is now a subsidiary of Compaq Computer. Tandem was
first to market with fault-tolerant computers.
The reason Sun has been successful in taking market share away from its rivals is simple. Unlike its competitors, which have been touting
"five nines" as the holy grail of computing availability, Sun already sells computers (Netra FTs) that provide greater than 99.999%
reliability and availability. Such performance translates to less than five minutes of both planned and unplanned downtime each year, at
two-thirds the cost.
There is a lot of growth in the demand for telecom equipment as a result of new players, such as competitive local exchange carriers
(CLECs), entering into the market," says Neil Knox, general manager and vice president of network systems at Sun Microsystems.
Sun had less than $350 million in telecom-related sales five years ago. That number has grown fourfold to $1.5 billion, equaling phone
company sales generated by rivals such as Tandem, Hewlett-Packard and International Business Machines.
Sun has also adopted some innovative marketing ploys to push its computers into the telecommunications space. For example, wireless
phone equipment maker Qualcomm merged Netra FT servers with its Q-Core transmission hardware. Lucent Technologies, on the other
hand, is taking Netras and bundling them right into its switches. Such deals mean that Sun sells more machines and has fewer marketing
costs.
Wall Street loves that combination. "Sun is gaining momentum in the telecommunications market as wire-line (also known as land-line)
and wireless players increasingly use its Netra FT servers to build out intelligent networks," says Amit Chopra, an analyst with N.Y.-based
investment bank Credit Suisse First Boston.
These Netra FT servers sell for more than $150,000 and carry a gross margin of greater than 52%. But that is not all. Along with these
Netra FT servers, Sun also sells $100,000-plus storage systems and the Enterprise 10,000, which handle customer
databases and other services.
The next-generation-network revolution provides Sun with a big opportunity to outpace its rivals in the telecom equipment space.
CSFB's Chopra points out that telecommunications-related sales will power Sun's total sales to $11.8 billion in fiscal 2000, up from this
year's projected $10 billion. Chopra forecasts the company's net income will hit $1.3 billion in fiscal 2000, up
from about $1.1 billion this year. It's no surprise then that Sun's stock has climbed 198% in the past 12 months to a split adjusted $60 a
share.
"The reason Sun has been so successful is that it is so committed to this market and is pushing the envelope," says Sanjiv Ahuja, president
of Morristown, N.J.-based Telcordia Technologies, formerly known as BellCore. Of course, it helps that Sun computers are extremely
stable, he points out. Telcordia makes the software that runs on different types of computers made by Sun and allows telephone companies
to offer features like call waiting and caller ID.
Telcordia is also part of Sun's plans to increase its stranglehold on the telecommunications business. The two companies are working
together to come up with a hardware and software platform for the "next-generation networks" (NGN) that will carry voice, data and video
over one wire. A current example of an NGN is Sprint's integrated on-demand (ION) network, which will give customers high-speed
Internet access and phone and video services over a single wire.
The NGN revolution provides Sun with a big opportunity to outpace its rivals in the telecom equipment space. This revolution has been
catalyzed by the shift of communication technology from circuit-switched to packet-switched networks.
Big investments in these future packet-based networks are currently under way as newer phone companies such as MCI WorldCom, Qwest
and Level 3 roll out high-speed IP-based networks. Sun wants to claim a large piece of the billions of dollars at stake. With Telcordia
developing the software, Sun providing the platform (servers and Java-based technologies) and equipment makers like Lucent and Cisco
Systems providing switches, Sun is looking to become the equipment supplier of
choice for every type of communication company. "Sun wants to be the lumberyard of choice for any type of service
provider--Internet services or telephony," says Sun's Anil Gadre, who knows that his company has a head start over rivals in this space.
The only thing that can stop Sun is Sun itself. Given McNealy's track record, don't bet against him.
____________________________________
Sun's Winning Storage Strategy By O. P. Malik
A trump card, according to Webster's, is a decisive overriding factor. For high-end server maker Sun Microsystems, the trump
card is the company's storage system business. Surprisingly, Sun is one of the biggest players in the high-end storage market, a quiet,
unnoted fact which is the way the marketing mavens in the Mountain View, Calif., company's headquarters like it.
Sun executives do not give the impression that the storage business is anything exceptional. Storage is not in founder Scott McNealy's
current stump speech about the enormous potential of Java programming language and network computers, the catch phrases of the
modern computing universe.
The truth is that each storage system Sun sells is paying for a significant portion of every one of these high-profile projects. So much so, the
storage division for Sun is the proverbial cash cow.
Glamorous no, but profitable, yes.
Sun is selling a lot of its high-end Enterprise 10000 servers, taking market share away from the likes of Digital Equipment and
Hewlett-Packard. So much so it cannot make enough to meet the high demand for these machines.
Each Enterprise 10000 retails for about $1.2 to $1.4 million. But the sale of every such computer drags along another $1 million in
storage array sales. According to Dataquest analyst Tom Lahive, Sun's gross margins on these storage devices could be anywhere between
35% and 40%.
Over the past three years, Sun has emerged as the largest player in the UNIX high-end storage device business. While these storage devices
do almost the same job as the hard drive on your laptop, the difference is the amount of data which can be stored in these devices. A few
terabytes. The high-end storage devices also known as arrays are like stacks of hard drives, one atop another in a shell.
According to Robert Gray, storage analyst with International Data Corp., Sun's 1996 revenues for the division were $1.32 billion, double
what it made in 1994 and $30 million more than IBM, which sold $1.29 billion worth of such devices.
Hewlett-Packard sold $1.13 billion worth of server-related storage devices, while EMC Corp. came in fourth with $688 million in sales.
The storage business is one of the fastest growing businesses for Sun. While the company's overall revenue grew 6% to $7.86 billion in
1996, the storage division saw revenue catapult by over 50%, according to IDC. Sun now controls 10% of the storage market.
Quite an achievement, since Sun did not have any array-based hardware systems on the market until 1994, and was losing sales to rivals
like EMC and IBM. So far the company has sold its products exclusively to the Sun customer base.
What if the company started selling these devices to users of other UNIX servers, from the likes of IBM and Hewlett-Packard? Some
analysts say that the revenues could go skyrocketing. Sun is taking on that challenge and is ready to go into the "third party" storage
business, which is targeted at the non-Sun machines.
Offense / Defense
The disk array market place is becoming separate from the server business, and the end users are willing to buy their servers from Sun, but
are thinking about buying storage from other vendors like HP," says Lahive. He points out that Sun's move might be "as much a defensive
move as an offensive maneuver."
Disk/Trend president James Porter is a little skeptical about the chances of Sun's making a go of it in the third-party sales business. He
thinks Sun might have a third-party strategy on paper, but feels the company lacks the sales and service infrastructure to go after this
market.
Sun's strategy is first to take away from third-party sales in its own Solaris environment and then sell to Windows NT and the various
versions of UNIX operating systems, with mainframe connectivity coming last.
Sun is not talking this challenge lightly and only recently announced that it intends to acquire the storage product arm of Fort Lauderdale,
Fla.-based Encore Computer Corp. for about $185 million. This move, according to some analysts is Sun's way of getting into the
lucrative high-end data center market.
Analysts say that Encore's Infinity SP product has the ability to share data between traditional mainframe and open-systems environments
by using a single set of data. This is a technology edge Sun wants, since other high-end storage vendors like EMC Corp. and Hitachi are
not expected to deliver until the end of 1997 at the earliest.
"The Encore acquisition does provide Sun a foot in the IBM space, since Sun's Enterprise 10000 servers are making inroads into the
mainframe space, and has done exceedingly well against IBM's SP/2 mainframe," says C.B. Lee, analyst at Sutro & Co., a San
Francisco-based brokerage firm.
Disks within Encore's Infinity SP can be accessed using software like NetWare, TCP/IP, or AppleShare eliminating the need for bulk file
transfers across the network or between storage environments. This product would give Sun six to nine months to establish a limited
market presence as a storage vendor in mixed mainframe and UNIX environments.
Following close on its heels would be a storage operating system software from Sun, written in Java, which would allow Sun storage
devices to be hooked up with other "brand name" servers as well. Lahive and Lee agree that Sun might be able to pull it off because of its
aggressive pricing tactics. "Sun has always been known as the price leader and they have the lowest-priced product in the market right
now," says Lahive.
Meanwhile expect more fireworks in this relatively quiet and little known business, as others like International Business Machines launch
their own cross-platform initiatives. |
