Original URL: http://www.theregister.co.uk/2006/07/21/hp_ae/

HP finally gets its own vision and can sell it too

'We learned'

By Ashlee Vance

Posted in Servers, 21st July 2006 22:08 GMT

Analysis It used to take HP about two days, ten executives and 20,000 words to near a coherent explanation of the Adaptive Enterprise. Thankfully, the company has put that ugly past behind it and can now describe its overarching strategy in a quick, palatable fashion.

The Adaptive Enterprise came to life under former CEO Carly Fiorina. At the most basic level, it stood as HP's catch all phrase to distinguish its sales approach from that of IBM, primarily, and also the likes of Sun Microsystems, Dell and EMC. Under the Adaptive Enterprise, customers were meant to sling around technology resources with ease. Applications would magically make their way to powerful servers when they sensed increased demand, storage systems would configure themselves into one, giant array, and robots would hover over your raised floor, blowing on boxes to keep them cool.

This vision - because that's all it was - culminated with the Utility Data Center package, which was a bunch of hardware and software wrapped in Fiorina's vague hokum.

HP has since scrapped the Utility Data Center idea and has something quite different to ship to customers these days.

"We learned," said Lin Nease, HP's director of technology strategy for servers and storage, during a meeting today with reporters at the company's headquarters.

Rather than selling customers a massive data center of the future, HP wants to ship smaller bits and pieces now, and then refine those components over time. The idea here is to add intelligence to your data center where appropriate with the hopes that you'll eventually craft a cheaper system that requires less space and less management.

Like any vendor, HP still coats its sales pitch with mounds of buzzwords and piles of ambiguity. But that's not to say there isn't something concrete lurking underneath the marketing goo.

The major premise that HP is working off comes from data that says the average customer spends about 90 per cent of its technology funds on maintenance, migrations, updates and other operational functions. That leaves only about 10 per cent of a company's budget to spend on new technology projects.

HP plans to improve that ratio by focusing on six areas - services, power and cooling, management, security, virtualization and automation. Spending too much time on HP's broad ambitions with each of these elements wouldn't do much good, so let's try and peg a couple of specifics.

Last month, HP rolled out its second take on blade servers in the form of the c-Class systems. Those in the know think HP got just about everything right with the new blades by giving customers a pretty dense case that has plenty of processor, component and networking options.

The c-Class blades also fit in well with HP's Adaptive Enterprise vision. You'll find, for example, some technology dubbed Virtual Connect that keeps track of the connections between servers and storage systems for administrators. If you start moving around blades or virtual servers on the blades, then the IP addresses, DNS and SAN (storage area network) data moves with the servers.

Beyond this, HP sells the sweet ProLiant Virtualization Management Software (VMS) package. This software gives administrators serious control over their virtual servers, including tools to move virtual servers around and to manage both VMware and Microsoft Virtual Server systems from one place. You may find a couple more bells and whistles by turning to new code from VMware or Cassatt, but HP has been shipping this package for months, and it's solid.

One could argue with relative ease that HP's virtualization tools are then even stronger on its Itanium-based servers running HP-UX.

We don't think HP has done enough to promote all that is available in the way of management software from the ProLiant and OpenView sides of the house. Entire start-ups have been built around similar management packages, and they garner tons of attention. It would be easy for HP to steal some of their ink.

The point being that HP has some very sophisticated gear available for purchase now - not five years from now. With a bit of courage and some smarts, you just might be able to replace a bunch of hot, fat servers with HP's blade gear and save on energy costs while getting an easier to manage overall package.

Many of you must be thinking that IBM can match HP's breadth of software and hardware products, and know that Sun comes pretty close as well. Thankfully, HP has decided not to waste peoples' time trying to debate this notion.

What HP now claims is that it offers more of a hands off approach when compared to IBM.

"With IBM, it is all about labor. It is about applying more people," said Olivier Helleboid, HP's VP of Adaptive Infrastructure.

HP argues that IBM's consultants are stuck trying to protect their salaries. They march into a customers' data center, take it over and then stay for life. HP, by contrast, wants to give customers tools so that they can help themselves. Of course, if this means paying HP's services team to figure out how to use the tools, then so be it. Wink.

Meanwhile, Sun has gone just too far with its ideas of putting up web sites where customers can rent CPU hours via credit card and of wanting to ship customers a limited set of gear, according to HP.

"We don't think selling CPU hours is a good business," said Russ Daniels, HP's CTO of software. "Just saying, 'We will offer you more CPUs' doesn't help us or the customer."

There's nothing really awe inspiring about what HP is pitching these days. In many ways, it's much the same virtualization, management, consolidation and service oriented architecture spiel that you can hear from any vendor.

But we like HP's decision to dial down the rhetoric and tackle this data center vision thing one piece at a time.

It has been a really long time since HP was able to convince us that it had a firm plan for helping customers. In the past, HP seemed more intent on defining its business strategy based on how IBM and Dell were defining their strategies. Now, as Daniels put it, HP has placed some bets and is willing to see if they're the right ones. Thankfully, these bets involve products and not just PowerPoint slides.

Of course, this could all just be the show HP dangles in front of the press on a Friday afternoon over cantaloupe and pasta salad. Is HP really helping you out? Let us know. ®