HP’s Darwin continues to evolve

While this month’s dismissal of former Chief Executive Officer Carly Fiorina has cast some doubt over the future of all of HP’s strategic plans, its technology and services offerings for big customers, key to doing battle with archrival IBM Corp., are likely to come under special scrutiny. Soon after IBM debuted its “on demand” computing vision in late 2002, Hewlett-Packard Co. (HP) joined the pack of vendors following IBM’s lead and rolled out its own, similar strategy for helping its customers build more agile and flexible technology infrastructures.

HP called its spin the “adaptive enterprise” strategy, and placed at its center a reference architecture, dubbed Darwin, intended as a road map for linking business needs to IT. Now, two years later, HP is making progress in transforming how it and its customers look at enterprise IT — even as HP’s executives struggle to explain the tangible components of Darwin and their adaptive enterprise vision.

The new HP CEO, when one is named, will have the chance to re-evaluate the company’s initiatives and ways of positioning itself in the market, including its adaptive enterprise efforts.

Still, company executives say they don’t anticipate disruption. “The management team and board are aligned on strategy,” said Nora Denzel, HP’s general manager of software and leader of adaptive enterprise. “The differences (with Fiorina) were on how to execute the plan.”

Denzel sees HP’s initial adaptive enterprise strategy as a three-to-five-year plan, on which it’s running exactly on target. She also says HP has had an easier time communicating the adaptive enterprise message than it’s been given credit for. “Adaptive enterprise is really resonating with customers. We’re having no problem communicating what it means,” she said. “We’ve never been thrown out of a company because they didn’t get what we were saying.”

Forrester Research Inc. analyst Frank Gillett says that HP’s board has indicated that it is committed to Darwin and the adaptive enterprise strategy even as it changes executive management. “If you’re going to the enterprise, you have to have a vision of what the future data center looks like and how it supports the business,” he said. “The existing management team has all bought into this. It’s not like Fiorina was imposing this on people against their will.”

HP launched Darwin in a blaze of hype in May 2003, on the one-year anniversary of its Compaq Computer Corp. acquisition. Since then, the architecture has faded from public view; while IBM executives rarely utter a sentence without invoking the company’s on-demand mantra, HP’s Darwin work has been a more internal affair. (It does still occasionally creep into the company’s public rhetoric. In her speech at Oracle Corp.’s most recent user conference, Fiorina commented on the similarity between HP’s Darwin architecture and Oracle’s software design.)

Still, Darwin remains an active, unifying strategy at HP, and the company recently revised the white paper and supporting documents it uses to explain its vision to potential customers and to its own sales and consulting staffers.

“It’s an implementation-agnostic reference architecture,” said Ram Appalaraju, vice president of HP’s adaptive enterprise program. “We give clients a model to help them think about building the management capacity to link their IT and business environment.”

The term “architecture” makes Darwin sound more technical and tangible than it is. HP executives sometimes refer to it as a “prescription” or an “approach” to designing IT systems. “We look upon the whole Darwin architecture as being about design rules to assist the CIO with the decision-making process,” Appalaraju said.

When HP enters into a large-scale customer engagement, the factors that must be considered can be wide ranging. Typically the work begins with a discussion of business processes and how they relate to technology, then progresses to higher-level design principles, and then finally to talk about specific technical architectures. “It is challenging sometimes to communicate effectively because the canvas is so large,” said Russ Daniels, an HP vice president and chief technology officer for software and adaptive enterprise.

“Darwin is our systematic approach to realizing the adaptive enterprise,” Daniels said. “We think if you don’t take that kind of comprehensive approach, then in fact your improvements will be relatively modest.”

Because Darwin is more about a modular philosophy than a concrete blueprint, it can be difficult for HP to pin down how the strategy influences its product and services plans. But Forrester’s Gillett said Darwin’s vagueness shouldn’t be mistaken for ineffectiveness.

“It’s conceptual. It’s not the sort of thing you can easily put in HP’s advertising, but it’s the linchpin of their adaptive enterprise strategy,” said Gillett, who specializes in enterprise IT infrastructure.

As HP’s thinking around Darwin evolves, the architecture has come to have less emphasis on defining specific components of the infrastructure — things like servers and storage and networking — and more emphasis on how IT itself has to be viewed as a service, Daniels said. This vision creeps into the product portfolio. For example, HP is developing resource allocation software, code-named Tycoon, that uses market mechanisms rather than business rules to decide which internal business units get priority access to the data center.

Tycoon awards users credits, which can then be used to bid for data center resources. When servers are in high demand, the purchase price goes up. Because prices can also drop during down times, departments are given a market-style incentive to spread out their computing jobs.

The software is being evaluated by CERN, the European Organization for Nuclear Research, which is considering it as a way to manage the mammoth computing needs of its upcoming Large Hadron Collider (LHC) particle accelerator, which will go live in 2007.

Though CERN has not committed to using the software in conjunction with the LHC project, researchers there are “interested in understanding if an auction-based approach makes sense inside science grids,” said Sverre Jarp, the chief technical officer of CERN’s openlab project, which is conducing the evaluation.

Daniels expects that Tycoon will be eventually delivered as an HP product, but it is an early-stage project not yet on the company’s road map.

Forrester’s Gillett sees HP and IBM at the vanguard of a move toward more flexible IT architectures. The change is inevitable, but so complex it will take years to spread through corporate IT departments, he said.

“I’ve only found a handful of large firms that are even thinking conceptually about this approach of rethinking the architecture of the data center, and so far all the ones I’ve talked to are sort of pursuing an inside path,” Gillett said. “Once firms face how complex this is, I think they’ll get more serious about it.”

One company staring down the complexity is Grand Rapids, Michigan-based office furniture products and services firm Steelcase Inc. Company Chief Information Officer John Dean is responsible for the IT needs of an organization with more than US$2 billion in annual revenue and 14,000 employees worldwide. He’s been with the company for more than 20 years, and in 2000 he watched the furniture industry fall off an economic cliff.

“We went immediately from a 50-year high to a 50-year low in capital spending,” Dean recalled. “You go through a shock like that, and agility and adjustable capacity become very important to you.”

Steelcase has been a long-time buyer of HP’s servers, desktops and laptops, and Dean took notice when the company began talking about its adaptive enterprise initiative. Last year, he signed on with HP’s services group for consultation work on assessing Steelcase’s infrastructure. “We believed there could be some lessons learned from what HP went through with (the Compaq) merger,” Dean said. “Steelcase is also a big accumulation of acquisitions.”

Working with HP, Dean determined that Steelcase needs to simplify its IT portfolio and evaluate its diverse business processes around the world to determine which really do require regional distinctions and which can be standardized. “The result was a blend of ‘we knew that’ to ‘we didn’t think of that’ — it adjusts your thinking in terms of your priorities,” he said of HP’s assessment work. “We didn’t have a strong point of view around standardization, not to the degree that we should have.”

Implementing technology is the easy part of organization transformation, Dean said. It’s identifying your business strategies, determining what actions are needed to achieve them, and making philosophical decisions about IT — does your organization regard it as an asset or an expense? — that’s tricky. HP understands that mindset, according to Dean “HP gets it,” he said. “That’s essential. I can’t move to a more modular architecture by myself, I can only move with my partners.”

(Scarlet Pruitt, in London, contributed to this report).

Related Download
Virtualization: For Victory Over IT Complexity Sponsor: HPE
Virtualization: For Victory Over IT Complexity
Download this white paper to learn how to effectively deploy virtualization and create your own high-performance infrastructures
Register Now