HP has come to a crossroads as a venerated but fallen-on-hard-times leader of the U.S. computer industry. It remains a trusted brand. Few companies enjoy higher customer confidence in their fundamental technology competence and objectivity in dealing with competing technologies, such as Windows versus Linux.

But trust has a second element that's currently running in short supply--the HP customer's sense that the company knows where it's going.

With its remarkable start-and-stop performance under the ill-fated Leo Apotheker, direction took a holiday and a new, uncomfortable- looking giant emerged, an HP that stumbled first in one direction, then the other. It was going to enter the tablet business, then, no, it was withdrawing before it incurred major losses.

It was reconsidering its PC business, then, never mind, of course it would sell PCs in order to remain competitive in servers. Its parts contracts were dependent on PC volumes. Why hadn't we thought of that before, HP spokesmen seemed to say.

Meanwhile, major customers were considering what they could do to restore certainty to their buying plans. A competitor's event, Dell World, was held in Austin, Texas, for the first time soon after HP's imbroglio; Michael Dell used the event to say he was staying in the PC business.

[ Learn more about an HP long shot: Project Moonshot, HP's attempt to bring low power servers into the data center. ]

What's an aging giant to do? When Oracle CEO Larry Ellison announced--without warning to HP--Oracle's end-of-life plans for HP's Itanium chip, it was like a brash young prizefighter trying to deliver a knockout punch to an aging champion. It turned out that there were no HP end-of-life plans for Itanium, thanks anyway for your thoughts, Mr. Ellison. A district court ordered the reinstatement of "a clear and unambiguous agreement" between the two that provided ongoing Oracle support. Nevertheless, confidence over HP's direction took another blow.

CEO Meg Whitman, a decisive chief executive of eBay during a 10-year stint there, replaced Apotheker at HP in September 2011.

Wednesday, Whitman told analysts and investors at the company's annual meeting that it was going to take much longer to turn around a "bloated" HP than she had anticipated. The revamp might continue through 2016.

Furthermore, profits per share for the year would fall in the $3.40-$3.60 range, well short of Wall Street's consensus level of $4.06. The statements brought immediate investor retribution. On a day when Apple staged another rise to $671.32, HP slid 13% to $14.91, according to Bloomberg.

In her Wednesday meeting, Whitman had to admit she and her team did little in their first year to arrest the decline. HP stock has lost 50% of its value in the last 12 months. She and top executives acknowledged efforts to apply quick fixes had failed and said the patient was going to need a longer convalescent period. Printers have been a good business for HP during the heyday of the PC, perhaps too good. In 2002, they constituted 40% of revenue and 95% of profits, Whitman explained.

The high margins allowed the company to paper over struggling business units and low-margin product lines as it attempted to diversify and firm up its business, an effort that's been only partly successful, as PC revenue dropped 10% over the previous year in the third quarter and printer revenue was off by 3%. Today, printers are 20% of revenue, though still 30% of profits. Whitman's reference to "bloat" probably included the 2,100 different laser printers that HP produces, many of them minor variations, but nevertheless discrepancies that cost HP efficiency in manufacturing and service.

The number of printer SKUs will be reduced by 30%; the number of PC SKUs by 25%, Whitman said. A reduction in force of 29,000 people, with 12,000 to be completed by Oct. 31, is in the works for a world-wide workforce of 349,600.

A Silver Lining In Cloud

On the more positive side of the ledger, one surprise was the amount of revenue that HP says it has derived from customer use of cloud computing. HP has assigned several data centers or parts of data centers to provide infrastructure-as-a-service offerings and introduced special services to usher customers into the brave new world of cloud. That strategy may be paying off, as HP says it projects $4 billion in cloud revenues this fiscal year ending Oct. 31. That would be a 39% increase over 2011.

HP's version of cloud computing got off to a shaky start during the Apotheker era. But it's also an area where HP's strengths regarding integrating servers, storage, and networking (and running them under one management system) may be producing greater results than expected. HP produced its own set of cloud management software to open its infrastructure-as-a-service. At the same time, some of the "cloud" revenue may come from HP re-labeling its well-established hosting service business. With definitions only loosely fixed between the two throughout the industry, some blurring of lines has occurred in many camps.