If HP can grow cloud services, it may also begin to realize some return on its $13.9 billion investment in Electronic Data Services in 2008. HP took an $8 billion write-down of that investment in its third quarter, a move that made up the bulk of its $8.9 billion loss in that period. That may have been a Whitman stage-setting maneuver that will allow HP in an upcoming quarter to say that its services business has resumed growth.

One small example of how HP is leading customers to cloud offerings: hand-holding by a customer service engineer is a feature of a new HP "proof of concept" cloud service. The engineer ensures a customer's initial attempt to run a workload in an HP cloud will succeed at an entry-level price. More cloud computing by that customer becomes likely to follow, according to Patricia Wilkey, global director, cloud services.

Project Moonshot: A Long Shot?

Another HP growth possibility is Project Moonshot. HP has been quicker than other hardware companies to implement a server design using the ARM processor, one from Austin-based Calxeda, right in Michael Dell's backyard. Cloud infrastructure at Google, Microsoft, and Amazon is uniformly based on Intel and AMD x86 chips. But HP is building data centers for a handful of large customers that want the small footprint servers that use 10% or less of the power of an x86 server. Each server does less work than an Intel counterpart, but more of them can be packed into an equivalent space, while still reducing cooling and other ancillary costs. ARM is the architecture typically powered by the weak batteries found in smartphones and other mobile devices.

A growing concern about new data centers contributing to global warming, as expressed by a much-debated Sept. 23 New York Times article, might end up spurring ARM server adoption, a field that HP owns at the moment.

Moonshot remains a long shot. Without a drastic change in data center user preferences, it will remain an experiment that hasn't paid off.

In other areas, HP faces a continued and steep uphill climb. It has established a software business that could be ranked as the sixth largest software company in the world, according to IDC. That business concentrates around network and system management capabilities represented by HP's one-time market leader OpenView, now called Network Management Center. But HP lacks a rapidly expanding engine inside its systems management portfolio that allows it to grow its business. It needs one to stave off the rapid expansion of VMware's vSphere 5 and Microsoft's System Center as they take on virtual machine management responsibilities.

At the Interop show in New York this week, HP has launched its strategy for moving into the software-defined networking space, a segment of the network market with a large potential for growth, as data centers continue to virtualize. But major revenue from the segment may still be years away.

Meanwhile, Whitman still has to convince analysts and investors, who were shaken by the dim economic news yesterday, that the company actually has a direction amidst these varied and uncertain opportunities. And she must convince customers of HP's sense of direction.

Find out the nine questions you must ask before migrating apps to the public cloud in the Cloud Ready? special issue of InformationWeek. Also in this issue: It's time to lay to rest two common myths of the cloud computing era. (Free registration required.)