Sun Microsystems, Inc.
Thoughts on Sun's Future

Format for Printing

Format for printing

Request Reprints


By Conehead
April 5, 2004

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

OK. I've now posted two long-winded posts on Sun and Linux and Sun and Microsoft. Both of which I've said are actually positive things for Sun. But what do I feel about Sun as a turnaround story at $5.00 /share?

Well the fundamental issue is that Sun is not a particularly diversified company. Frankly, Sun was the only profitable enterprise hardware company for a long time. HP loses money in its server division, making most of its profit from printer ink. IBM's server division is also unprofitable, making up for it in consulting. I'm not 100% certain about niche players like Fujitsu, but I don't think that they are profitable either. Intel's server chip, Itanium, has been enormously unprofitable and AMD hasn't made any money in its server division either.

So, while the rest of the world can "wait and see" what happens in the enterprise space, or even use it as a loss leader like IBM, Sun is forced to immediately find a way to make the enterprise space profitable. Sun has to adjust its business model while the "plane is in the air".

The following is speculation on my part and is not, by any stretch of imagination, an official Sun statement. This is why I try to stay pseudo-anonymous. (Although I wonder if anyone has figured out my real name.)

Sun's problem right now is (without doubt) average unit sales price. Sun is continuing to ship more and more servers. But, Sun's best profit margins lie in the mid to high end (24-72 CPU) space. Too many of Sun's sales have been in the low to mid (4-12 CPU) space. Sun's current business model and sales force is geared around the half million-dollar server sale with 30% margins. In other words, a very "high touch" sales model. But that ERP system that once cost a $500,000 dollars can now be run on a $100,000 V880 with much thinner margins.

I believe that Sun sees it high-end server business as a legacy business. A very important legacy business, but not one that will enjoy significant growth in the future. I think that they see high-end systems as the mainframes of the 21st century. They will be the systems that companies bet their business on. They will continue to add new features that enable powerful consolidation and high availability features. They will continue to see single digit growth as smaller systems are consolidated onto "big iron". But they won't be able to sustain double-digit growth for extended periods of time.

So, I think that Sun is trying to develop a three-pronged attack to develop a new and profitable business model.

1. Reduce costs. This means several things. It means outsourcing a lot of the company. It also means trying to make its sales force a lot leaner. It also means using commodity parts whenever possible. This even getting past the "Not Invented Here" syndrome to adopt technologies like AMD.

2. Increase attach rates of higher margin products. Sun has famously struggled to increase storage attach rates. They are now fighting the software war as well. The goal obviously being to "upsell" customers into higher revenue, higher margin, and higher loyalty product lines. (I once heard the expression "IT managers may love their hardware vendors, but they are married to their software vendors." )

3. Use the freedom of being an open systems company to develop solutions that create either new markets or quantum leaps for existing markets. Sun is pretty unique. It is largely independent from the Intel monopoly (it sells some x86 servers, but not so many as to be dependent on the revenue), and completely independent from the Windows monopoly. As such, it can do things that other enterprise vendors cannot. Such as market a Linux desktop. Or market an Office competitor. Or create a server chip like Niagara that forgoes single threaded performance for high thread level parallelism. Obviously new and innovative products are going to carry higher margins that again can leverage this "high touch" enterprise sales force to gain penetration.

So, what do I think about the viability of this strategy?

It's always been about execution.

This board has largely been critical that Sun has been an "academic institution" focused on the technology, but not the business. I find this untrue. There is no one at Sun these days that isn't focused on generating revenue and profit. Unfortunately, I still find a lack of focus. In the 2004 fiscal year, I will have undergone three major reorganizations, and two major layoffs, and one minor "redeployment" layoff. Needless to say, this isn't particularly conducive to either productivity or loyalty. I don't think that the technology is the distraction from the revenue; I think that it is the politics that is the distraction.

LostInColorado states that there are more managers involved in his project than there are developers. Similarly, I have more "overlays" and managers in my group than sales executives and sales engineers. These are generally people who have enough friends and seniority to avoid getting laid off, but what have little to actually contribute to Sun.

I'm hoping (perhaps too optimistically) that Jonathan Schwartz's promotion to President/COO will help with this. Jonathan is very focused. He also isn't "old school Sun". I'm hoping that he can cut through the deadwood and Sun and execute on the strategies. Scott McNealy has always been forward thinking. But I don't think he was able to pick good executives and was unwilling to "clean house" until it was too late.

Unfortunately, I think there are some rough times ahead for Sun. Even if Sun executes effectively on the strategies above, Sun is likely to have a tough time adapting. For example, imagine that Sun's Niagara chip ships and is truly able to offer a leap ahead in throughput computing. Just to put a number behind it, lets say that Niagara chips have four times the throughput of the equivalent AMD or POWER processor. So, hypothetically, imagine that this increases Sun's market share from 30% to 45% in the UNIX+Linux space. Even with the increased customers (and assumably increased market size), these small Niagara systems will be putting increased pressure on the higher end midrange systems. If a Niagara blade can replace a V440, that's not exactly a pretty picture.

Similarly, I find lots of people at Sun saying the "future of Sun is software". Well at $100/employee/year (including support/OS/education) it would take some pretty impressive deals to justify a company the size of Sun. BEA generates about a billion dollars a year on 3000 employees. Even assuming that Sun can sell 10 million seats of JES to create a software division as big as BEA, Sun would still only need 10% of its employees. (Again assuming BEA as a benchmark.)

The bottom line is that I think Sun is a company that is going to have to completely reinvent itself. Which is going to be pretty painful for us employees. But this transformation is certainly achievable; Sun has re-invented itself many times. IBM reinvented itself from a hardware company to a services company. If Zander can transform Sun from a workstation company to a server company, I certainly believe that Schwartz can transform Sun from a server company to a "systems" company.

The big question is trying to guess what Sun will look like on the other side of the transformation. At $5 a share it may be worth the bet. Frankly the worst case will still involve a long and sustainable legacy platform. There will be massive layoffs in this scenario, but there will be plenty of generated cash to give back to investors. On the other hand, the best-case scenario involves a completely transformed company. Creating a software division as successful as BEA would add nearly $2 a share to the SUNW ticker. Creating a storage division even half as successful as EMC would add $5 a share to the ticker. A Solaris x86 strategy win might win $1.5 to the share (based on RHAT price). Increasing margin 10% from reduced costs would likely double the share price, at least.

I think the question is one of timing, however. Is this the bottom? Is this the best time to buy, or would it be best to wait for signs of recovery/turnaround? I don't know. At least not yet. I think I may go back and study the history of some comparables.


Become a Complete Fool
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.