View from the Hill - #3

Should Sun Microsystems make its own brand of "Intel Inside®" PC's?

It could become a successful PC supplier a lot more easily than you might think... But does this make sense from a strategic business point of view?

Apple has been fighting the desktop war in its own way, and it looks like it will retain its distinctive alternative niche. Sun has never been successful in the desktop market but has many attributes which could enable it to successfully step into the Intel Inside® desktop arena. This article analyzes why Sun could be a powerful PC player, and why they will probably decide not to go down this road for another couple of years.

Some people at this stage, especially if they work for Sun, might say "what about the NC?". To which I would reply, take one home, give it to the kids for Christmas and see what they say. My guess is that they will find it only marginally more useful for their computing needs than your microwave oven, and a lot less attractive than a PC, a MAC or a Playstation. Warning! Don't try this experiment in real life, because the outcome could be dangerous.

I knew that Sun had lost their sense of perspective on the desktop when they introduced the JavaStation in 1996. Sun is good at hype, but this was the most hype I'd seen about a dumb terminal. OK, it was important, if you were a user of dumb terminals because VT100 emulation (an old DEC standard) made it difficult to implement the kind of windowing that people saw on their PC's at home. Hell, even a mainframe environment doesn't have to be boring.

There is a market for terminals, just as there is a market for railway rolling stock. But when it's your own money and you are looking for your first car what do you buy? Most of us don't rush out to buy a train. We buy a car, and work our way up the brands and pricing curve depending on our aspirations and life style. Maybe we'll buy trucks for our business, or even use railway freight services. But the NC is a large scale solution which doesn't scale down. That's one reason why it's never been successful. The other is, that as a computing platform, compared to a PC, it's not even on the same planet, let alone in the same race.

OK so why am I saying that Sun Microsystems could become a world class PC player, if it chose to do this. And why haven't they done so before?

Sun does know rather more about the insides of a PC than most of the companies which currently make them.

• They introduced their first Intel inside computer in 1987, the 386i. This was an underpowered workstation, but if you ordered it with a monochrome monitor, and used it as a SunOS databaserve server it cost a small fraction of the price of a competing VAX, and could run for more 6 months at a time between having to hit the reset button. A feat which many modern PC's running other software would find difficult to match even today.

• Sun also has a decade of experience with a variety of add-in cards, like the SunPC which has provided emulation of the Intel PC environment to early SBus, and now PCI based SPARC workstations.

• Sun also has produced one of the most powerful multi-user operationg systems which run on Intel boxes, namely Solaris.

If you add to these undoubted technical capabilities, their manufacturing strength:- the fact that they can make money out of monitors and NC's which have a lower price point than a PC. And their undoubted marketing abilities:- Sun have been marketing via the web directly to end-users longer than most of their competitors in the PC business. The web paradigm which scares most PC companies (except Dell) is already part of Sun's toolkit.

As we know, almost anyone can build a PC, and Sun has many attributes which would make them a perfect PC supplier. So why don't they do it. Let's look first of all, at why they haven't done it earlier.

In the period 1987 to 1998 Sun positively benefitted from not making PC's. Not making PC's gave them the following competitive advantages.

• focus - all their sales people could focus on one technology - SPARC, and one operating system - Solaris. Sales people are only human, and if your company sells a lot of different technologies, you will find it difficult to understand more than a little about any of them. Sun's salesforce were therefore able to be more competent about the products they were selling, because they had a simpler product line.

• evangelism - Sun's marketing messages in establishing SPARC as a strategic computer architecture were simplified by the fact it didn't sell anything else.

• simpler relationship with Microsoft . Sun and Microsoft are competing suppliers of operating systems in the server market, and they have never seemed to like each other very much. Therefore a Sun PC strategy would have been complicated by the need to buy a strategic component (the OS) from an unloved rival.

• profitability - the PC market had lower profit margins than the multiuser server market. So while Sun was establishing its dominance in servers, it was actually better off ignoring the PC,

• channel conflict - most resellers sell PC's. The fact that Sun didn't make PC's simplified its interface with the channel, which was strategically important in its business development.

If we look at Sun's options in 2,000 Sun has already established itself firmly as the #1 supplier of choice in the fast ecommerce server market. Profit margins in this segment are good, and the prospects for growth look solid for a least a couple more years. So Sun can afford to ignore the desktop for a little while longer, because from a business point of view, it currently has more attractive options.

However, what will happen is that a generation of MIS managers will come up the ranks of their organizations who have only ever known Intel architecture PC's and servers. Whether they are running Solaris on multiprocessor Intel servers, or Linux, or NT, the mid range will be squeezed hard and while SPARC will remain competitive in this area, the brand awareness of SPARC versus Intel could prove to be more important in that market space than any other consideration. That's why it's important for Sun to brand itself and its technologies strongly today for tomorrow's markets.

The channel conflict idea is still an important factor. The most successful PC companies are abandoning the their channels and selling direct. So implementing a successful PC strategy could put jeopardise a channel model which has worked very well for Sun in the past.

My guess is that Sun will choose not to become a PC manufacturer in the short term. They have much to gain and much to lose whichever way they go. But unlike many other companies, even the option of doing nothing new in the PC market could be quite attractive option for Sun because of the strong growth momentum in most of their current markets.

SPARC(R) is a registered trademark of SPARC International, Inc. SPARC PRODUCT DIRECTORY(SM) is a service mark of SPARC International, Inc used under license by ACSL. Products using the SPARC trademarks are based on an architecture developed by Sun Microsystems, Inc.