|A lot of people (me
included) missed the start of the 2000-2003 IT recession... |
Intel and PC makers first, and didn't seem to start affecting the Sun and
storage markets till about 5 months later. Sun may have brought itself extra
time by stuffing the channel with servers that wouldn't be needed, as was
reported to me later by a distributor which pulled out of the Sun market. But
that's an old story. I'm interested in what's happening right now.
just as likely that a lot of people will miss the start of the recovery
in the Sun market too.
Is there a recovery? - You may ask
Yes, well I'm just as surprised as you are, but that's
what's predicted by pageviews in the
SPARC Product Directory in
March which were higher than at the same time in 2001, 2002 and 2003. And our
SPARC readership is 22% up on a year ago too. I have to say this is not what I
expected. But then the market knows best, and here are the reasons why.
has been a recovery going on in the PC and non SPARC server market for 8 months.
Also a huge recovery in the storage market with dozens of manufacturers
reporting double digit revenue growth. (The storage market will grow faster than
the rest of the computer market, because the storage market is going to replace
a lot of servers. But that's another story.) Mysteriously the recovery in
server revenues reported by IBM, Dell and HP last year seem to have passed Sun
by. Sun users held back probably because they were wondering if Sun itself had a
future, but also because they weren't being offered the kinds of performance
enhancements that were being seen in the Intel architecture market. So why rush
to buy another slow box from a company which might not be around to support it?
Sun's recent deal with Microsoft has bought them the cash to ride out another 6
- 9 months of time to make their plans work. That's more than long enough for
Sun to start getting the positive benefits from their new SPARC IV servers.
Sun's biggest institutional customers will be brushing off the cobwebs from
their checkbooks and creating the biggest surge of orders that Sun has seen
since 1999. Sun's technology deal with Fujitsu (though not confirmed at this
time) also provides an assured stream of high end SPARC processors and servers -
with shared development costs and therefore less contribution to losses for Sun.
Most Sun VARs will miss the early signs of the recovery because they
have dug into their trenches so deep that they are invisible to potential
customers, and they have learned to mistrust the optimistic hype which flows
ever on down from Sun. So even when some of that hype is true, they'll ignore
But what's important is not what Sun is doing or saying, it's what
their customers are thinking and doing. They're always the ones who lead
the market, whatever the direction. Analysts interpret long after. ...ACSL (publisher)
article:- Storage Winners
and Losers from the 2000-2003 IT Recession
predicted in this article Sun reported year on year revenue growth for the next
2 quarters following publication. (Quarters ended June and September 2004.)
telco recovery will increase add icing on the cake
recession and restructuring in the telecoms market in recent years meant that
the industry had over capacity in servers. Historically the telco market was a
major user of SPARC systems. When the telco market recovers at the tail end of
2004 or early 2005 that will add more spring into the SPARC market's bounce
Nibble:- The Sun Microsoft Settlement
you cut away all the rhetoric about this deal the most important thing from the
viewpoint of SPARC readers - is that the $1.6 billion nett which Sun will get
from Microsoft buys another 6 to 9 months of life at Sun's current burn rate. In
that time a lot of things can happen.
Most significantly Sun's new
SPARC 4 servers will hit a market which is desperate for higher performance, and
moreover, because of the US economic recovery can actually afford to buy those
Provided that Sun does a good packaging deal in its top of
the line servers to integrate performance accelerators (such as wire speed
Fibre-Channel and TCP IP Offload host bus adapters, and fast disks - maybe even
solid state disks) the performance figures should be very competitive.
Intel which has to wait ages before Microsoft tweaks its operating systems to
use new hardware features, Sun as usual, has been designing the new version of
Solaris to do that all along the way. That's one reason why new SPARC processors
always look good - they benefit from simultaneous hardware and software
enhancements at the time they are announced.
For Microsoft the
settlement is small change, and may improve its legal battle with the European
Union. Its battle with Sun is old news. Microsoft isn't fighting that war any
more. That was World War I. Microsoft's current war is in the storage space.
When data storage systems became operating system agnostic that threatened to
make the server OS irrelevant. Microsoft reacted by providing a version of its
server software which actually runs on storage appliances. That way it wins
Microsoft lost the war to power cell phones (Linux
won that one) but the server storage war is more visible and has been getting a
lot more attention recently.
Back to the deal's effects on Sun... Will
we see a spate of new Sun AMD powered servers which are Windows compatible?
Although the settlement opens the door to that, nobody really cares,
and that's not the way that Sun will survive.
Sun will live or die by
its success as keeping SPARC servers running as competitive alternatives to
Intel Architecture systems. If IBM could do it with their mainframes than Sun
still has a fighting chance, and now it's got a much longer breathing space in
which to make that happen.