POST OF THE DAY
Sun Microsystems, Inc.
In Reply To:
Sun is Bashing the Wrong Enemy

Format for Printing

Format for printing

Request Reprints

Reuse/Reprint

By AngryOmar
July 14, 2003

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!

Okay, let's get a few facts straight here:

Chris:
In general Sun spends more time attacking IBM than Microsoft (that's my impression anyway). The pop press love to hype up Sun's Microsoft bashing to the max..

SunStruck:
While you may be right, they have been losing the battle gradually over the last decade.


Let's examine the last decade:

* From fiscal year 1993 to 2002 Sun's revenue roughly tripled from $4.3 billion to $12.5 billion (for comparison, Intel's also about tripled from $8.8 to $26.7 billion.)

* Sun went from being non-existent in the server market in 1993, to $6-6.5 billion and 13-15% market share (as of CY2002, IDC/Gartner data.)

* Solaris was released, and has grown to become the most popular version of Unix by a wide margin (margin much smaller if you count Linux, but still there.)

* Java was released, and in 7 years has become one of the most frequently used programming languages, and one of the most frequently targeted platforms. According to the last market research I've seen, Java is used by more developers than the combined C/C++, though only about half as many use Java as their primary language.

There have been many things that did not go so well. Java has not become successful on desktop PCs (yet), Sun and SPARC International were not successful in establishing SPARC as a popular standard architecture (the clone market never grew much, we'll get to this later.) But with the exception of the past two years, Sun has been extremely successful in the past decade. The facts make this pretty much indisputable.

If it was not for Linux and it's rescue of Java, Sun would be an insignificant competitor right now IMO

I'm sorry, I don't think that Linux has rescued Java, and it has had nothing to do with Sun's successes in the last decade. If anything, the internet has had more to do with Sun's success than any other external factor.

While Linux can also rescue the UltraSparc and help build out it's developer base, Sun has not taken advantage of this.

No. If Windows NT, a much more popular OS than Linux could not rescue Alpha, than Linux cannot help UltraSparc, which I don't believe needs rescuing anyway.

Sun already sells UltraSparc outside of Sun. The problem is Sun also competes against the customers of UltraSparc, which is why there aren't many computers using it, except from Sun. If you want to see, have a look on http://www.sparc.org . Making UltraSparc widely successful outside of Sun would probably mean exiting the market for computer systems, which would most likely be suicidal for Sun at this point. It would take at least this much to get IBM, HP, or Dell to use UltraSparc (which along with Sun and Fujitsu, pretty much ARE the server market.) Unless you are suggesting UltraSparc for personal computers, which is even more unlikely than Itanium for personal computers for all the same reasons: too expensive, no applications, no need for a new architecture.

Let's put the argument this way: If AMD or Intel were to develop a new chip architecture, then do you think it would be wise for them to develop their own os for it and market it vertically?

If their OS was already extremely successful in a particular market, had large ISV support, and generated lots of revenue then it would be stupid to stop developing it. This doesn't mean that other options can't also be pursued.

Hmm, sound a lot like Steve Jobs. Let's try and beat the horizontal guys with innovation. Obviously this hasn't worked for Apple and it never will.

In Apple's case, with regards to personal computers, I think you are partially right, but you're generalizing and over simplifying the situation (first of all, Apple don't design their own processors.) The IBM clone market (err, "horizontal guys") created a few situations:
1) Intense price competition between vendors, who had no other way to differentiate their products
2) leading to lower prices
3) leading to increased IBM-PC volume
4) leading to more ISV support
5) goto 3

I think that in addition to this, Apple kept their prices high to try and maintain their margins, which helped item #3 and accelerated the cycle (though I could be wrong.) Apple did eventually try to create a clone market, but this was not successful.

So how is the situation different in Sun's case?

First of all, Sun is being aggressive in pricing low-end systems. That has been demonstrated repeatedly on this board.

Second, there are currently important differentiators between 64-bit RISC computers and x86 computers in the market where they are sold, although people are figuring out ways to use PCs for tasks that normally require large servers ( including Sun.) By contrast, in the market for PCs Apple computers did not add enough value or differentiation to justify their cost for most users. Furthermore I think more applications were available on Windows, which made Macs less valuable. Today if you need a 64-bit server Sun is the dominant company, and the applications are all available for SPARC/Solaris.

With regards to the throughput computing strategy, the idea is to beat the price/performance of x86 based machines by targeting a different aspect of server performance. This is about designing for the attributes that are valued by the market. Our contention is that the server market values throughput more than single-threaded performance. This is mostly a result of the types of applications that are run on servers (web servers, databases, etc.), which have emerged relatively recently.

The typical Innovator's Dilemma scenario is that an "inferior" product is sold into a lower market, until the volume of the lower market enables a price/performance that is better than products in higher markets (where performance just means some attribute that is valued by the market.) Sun's belief is that the attribute of value is throughput, and furthermore that this is not valued in the lower market (PCs.) If successful, this strategy is very clever because it takes away the volume advantage of the lower market.

But whether or not this strategy turns out to be right may not matter. This is not the only thing that Sun is betting on. Notice the increased attention to software with things like project Orion, Madhatter, Java, and Solaris x86. As was mentioned, Sun has begun shipping x86-based servers, although whether or not this is a market that we want to or should be in is debatable (x86 software is a different story.) It is possible that SPARC and Solaris may become niche products in the larger server market, but this would take a LONG time. Even mainframes still generate about $5 billion in revenue for IBM (most of it profit, probably.)

In any case, from my perspective Sun is following strategies that will probably lead to continued success. They are different from what every other computer company is doing, but that's always been the case.

AngryOmar

P.S. For further reading, I'd suggest:

[1] High Noon, Scott McNealy and the Rise of Sun Microsystems. Karen Southwick. Surprisingly unbiased history of Sun, and despite the title doesn't talk much about Scott.
[2] Sunburst, The Ascent of Sun Microsystems. Mark Hall and John Barry. Older book on the history of Sun, written in 1991. Excellent to get a perspective of Sun and the computer industry from a pre-internet point of view (especially for youngsters like me.) Most notable for barely mentioning Microsoft, and not mentioning the word "server" at ALL! Might have to get it used.
[3] The Innovator's Dilemma. Clayton M. Christensen. Essential reading, enough said. Look it up on Amazon if you want to know more.


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.