The Boston HP Party

The week before last HP held a conference in Boston in order to share the love with the IT analyst community, so I flew Jet Blue from Austin just for the party. HP is gung-ho at the moment. I guess it has every right to be. The era of Carly-You're-So-Vain-Fiorina is clearly over, and since the advent of David Hurd, HP has been turning in positive financial results—almost from the get go. Indeed, HP's market value has grown by about 62 percent, under Hurd—which amounts to $32 billion. Not too shabby.

From where I'm sitting it looks as though it has been achieved simply by a focus on execution. There's nothing particularly dramatic going on at HP, except that Hurd appears to be fixing what was broken. HP was in the difficult position of being lodged in between IBM and Dell in the hardware market and it still is, but it doesn't look so uncomfortable any more. While IBM was gradually moving out of the PC business (by engineering a Chinese take-away), HP continued to go head-to head with Dell both with PCs and with low-end servers. In an unexpected turnaround, HP is now stealing market share from Dell, both in the PC and server market.

Nevertheless, there is no denying that HP's primary competitor is IBM. At the turn of the millennium, or some time around then, IBM started to drive forward with a focus on its new “On Demand” vision. It was a convincing push because, technologically, the flexibility that is at the heart of the On Demand message was and is what the IT world needs. HP's response was to come up with a similar vision—the Adaptive Enterprise. (Truth to tell, both companies were borrowing a little from Sun Microsystems, which was first with this type of message). The point is that this focus genuinely does put HP toe-to-toe to IBM, with Dell still trying to work out whether it matters.

So, how is HP doing?

HP ESS

ESS stands for Enterprise Servers and Storage. HP has three active server families; The Proliant Servers (which sell like hot cakes), The Integrity Servers (based on the Itanium chip) and the Non-Stop Integrity Servers (which is what the old Tandem business has morphed into). The Non-Stop servers keep selling (HP does about $1 billion revenue with these) and depending on how HP manages this business, it may undergo the same revival that IBM's mainframe is seeing.

A question mark still hangs over Itanium and it may be a year or two before the situation becomes clear. Itanium is HP's replacement for the remarkably successful PA-RISC chip and the might-have-been-if-only Alpha chip—it's a collaboration with Intel that has always run late and usually disappointed. Itanium is the foundation of HP's Integrity servers.

For the past three or four years there has been speculation that Itanium would never make it, but it made it into the Integrity server and HP is selling these servers. About 7000 ISV's have now ported their software and HP has probably got sufficient momentum here to ensure that Itanium has a future. The impression I got is that HP will continue to invest in Itanium no matter what. There is no plan B.

Will Itanium ever really make it? It's still too early to say, but it's very late to be too early to say.

HP Consultancy & Services

The services side of HP's business is now at about $15 billion and growing. HP is doing well here, mostly on the back of its own technology. Its consultancy business is still only a fraction of IBM's, but it becomes more significant every year. While IBM has got to the point where growing its consultancy arm is difficult, HP has not. HP, by the way, specializes in building data centers. And it has technology for keeping them cool.

Having run into the cooling issue just a week before, I was not too stunned by some of the facts that HP came up with in a session it gave on “staying cool”. But nevertheless they were still a little surprising.

Fact 1: The cost of electricity to run a server for 3 years is roughly the same as the cost of the server. (This is an average figure. of course).

Fact 2: For every dollar spent on powering the computer 2 dollars needs to be spent on keeping it cool.

Fact 3: Instances are now occurring of data centers going out of action because of cooling system failure.

Naturally HP is presenting such information because it has an answer. It believes can save data centers big money.

HP Software

HP's software division now has revenues of around $1 billion (taking OpenView & OpenCall together). It has been growing organically at an average of 15% (against a market average is 10%), with OpenView itself growing at 20% last year.

HP has long been a graveyard for good software—it could invent good software and/or acquire good software but consistently failed to sell it. That problem may now have been solved, because HP's software division is now large, growing and in profit. It qualifies as a serious business in my view. Anyway, HP needs OpenView to continue to be successful, if it is going to deliver on its Adaptive Enterprise.

HP announced that it intends to make acquisitions, including possibly a very large acquisition, in order to grow the software side. It denied having any ambitions to extend OpenView into the IBM mainframe market so, in theory, BMC is not on its list of possible buys.

One HP OpenView customer story (from the customer's mouth) told of money saved in implementing OpenView in an end-to-end way, replacing many point-solutions and reducing the number of infrastructure vendors the organization had to deal with by two thirds. The customer claimed that the savings were in millions (it was a large organization). This, I believe, is the big point in the infrastructure management space. Whether you go with CA, HP or IBM/Tivoli probably doesn't matter as much as replacing point solutions that prevent end-to-end coherence. Infrastructure management is tough and getting tougher. The big fish are eating the small ones.