Sun … dead, dead, dead?
Email This Post
The Tuesday San
Francisco Chronicle business section (an oxymoronic term if ever there was
one) radiates a bright, Sunny orange hue as a backdrop to a story that
glosses over Sun and their "daring comeback bid." The article echoes
the blog of
Sun’s pony tailed Comeback King, Jonathan Schwartz proclaiming "The
proprietary and expensive moniker is now dead. Dead, dead, dead."
Deny, deny, deny.
You might think that I’m preparing to bash Sun Micro yet again.
Well I am, but that comes later on down the page. I cannot pass the
opportunity to lambaste the Chronicle, a "news" organization not hindered by
journalism, for their angle on this piece. Naturally, they want to paint
any neighboring Silicon Valley advertiser in radiant colors, but even the
Chronicle had to stretch facts beyond their normal capacity for truth bending
(with only Pravda being better equipped at that unsavory art).
The article’s opening gambit is based on a recent Gartner report that (based
on server revenues ) puts Sun ahead of Dell and into 3rd place behind the
neck-and-neck pair of HP and IBM (and even then with less than 1/2 their
individual server revenues). By wrapping this marginal news item with
speculation concerning Sun retaining their loyal customers ( proof point? )
while reaching out to new customers ( proof point? ), the Chronicle positions Sun
as a revitalized success story that nobody knows about. They go as far as
saying "The company needs a positive buzz to reestablish itself as a top player
in a rapidly growing market for business information technology". Though
at the tail of the article the Chronicle alluded to Sun’s continuing fiscal
anemia, little was said about the lack of homogeneality concerning Sun’s alleged
strategy, their negative earnings, or why anyone thinks the light in the sky is
not Sun in the early phases of going nova.
Let me first berate the analyst groups over the ongoing inequity of using
revenue number as a basis for measuring market share. IDC and Gartner
obfuscate imbalances caused by relative pricing of large, small, and commodity
servers. They ignore market share as a factor of rate-of-adoption, which
would be measured in units shipped and not dollars charged. Finally, using
revenue as a basis says nothing about the long-term prospects of a vendor,
especially those hemmoraging cash. To be fair, if you spend enough money
with them, you can get unit shipment data, but that is never what they use to
generate headlines.
For the sake of investigation (a foreign word in the San Francisco
Chronicle lexicon), I grabbed annual financials for Sun and then Dell (who IDC
and Gartner claim has slipped to forth place). Sadly these numbers do not
reveal server-specific revenue or units shipped by Sun or Dell (more on that
momentarily), but they do show the top- and bottom-line effects of core
strategy. Dell has gained top-line revenues for each of the last five
years, while Sun has bobbled up and down with only a recent short spike.
Dell has had positive earnings in the billions of dollars, while sun has five
straight years of losses, with the last approaching a cool billion. (if
any of these numbers are off, blame Fidelity Investments).
The real kicker is the burn rate. Unless Microsoft Excel lied to me
(which is not unthinkable), Sun has 1.5 years of life left at best given their
current loss rate, and 1.2 years if you average their last five annual loss
rates against current net assets.
From a strategy standpoint, Sun is in a bind, though they are slowly finding
a path through the woods. That path might lead to a tar pit, but that
remains to be seen. At very least Sun wants people to believe that their
"proprietary moniker" is dead.
I’ll believe that one I see both Novell and Red Hat with a SPARC-based Linux
on their price list. Sun has augmented their product line up, and tried to
seed the market with free Solaris as a way to ease high-end server sales, but
their current solution set still reeks of selling in (with barely competitive
low-end offerings) to later sell up into … proprietary hardware.
IBM still sells proprietary hardware, namely mainframes. But in the
face of growing competition from commodity systems, they discovered how to make
mainframes part of the overall solution set, both as an expandable hub to
commodity forests surrounding that castle, and by sporting Linux on the
mainframe in logical partitions nearly a decade ago. Sun has awakened to
the fact that Solaris and SPARC are destined to be only a part of any solution
set, and don’t seem ready to find ways of being unique and competative in the
now dominate world of commodity computing.
The back story here is that HP has gotten downright feisty of late (is Mark
Hurd passing
Peruvion Marcing Powder around the HP HQ? ). While trolling for data I
discovered that
HP is not satisfied with analyst use of revenue as a market share indicator
either. In pages boldly titled "The Real Story about Server Market
Share", HP has claim to the #1 position based on units shipped, with an alleged
server ship rate six times larger than Sun. The telling part (if HP
calculators are still to be trusted) is that Linux on x86 is shipping 81 times
as many servers as Solaris on x86 shipments. This is perfectly believable,
and given the now dominent demand for commodity servers, shows the
Sisyphus-grade
task that Sun cannot likely accomplish in 1.5 years.
One last aside too comical for commentary (well, OK, I’ll make at least one
snide remark). HP slipped into abject sillines in their "real story" by
noting they are the #1 shipper of Itanium servers, holding 62% of that market.
I believe this means they shipped 62 Itanium servers last year.
