All of these projects face technical and market hurdles which could prevent them from becoming successful commercial ventures. Nevertheless the variety, depth and vitality of the work merits attention.

Handheld drivers

The projects stem from a belief that today's computers — including high-performance servers — are rapidly headed for commodity status, and that tomorrow's computer industry will be driven by a diverse array of handheld gadgets.

"I think we are finished in the office, and consumer systems mark the next big opportunity," said Dick Lampman, vice president of research for HP and director of HP Labs.

Lampman, an energetic, 30-year veteran of HP, speaks with great enthusiasm about the projects and the lively spirit at the Labs these days, including the IA-64 microprocessor architecture he helped define as part of an assignment in December 1988.

"I still believe that despite a few year's delay in launching Itanium, the server world is going to change," said Lampman. "High volume technologies will run from top to bottom in general-purpose computing. We are in a convergence cycle around high volume models in general-purpose computing. The real renaissance is in a coming generation of embedded non-general-purpose platforms."

Indeed, as HP was co-developing the Itanium architecture with Intel Corp., the company was pointing its research compass toward a horizon of handheld systems using diverse, optimized architectures. "The whole theme is that new devices will drive the future," Lampman said. "All this started in the early '90s."

Lampman is upbeat about the HP/Compaq merger, which is to be expected from the man who will run the combined companies' research operations. Some researchers likely will be laid off as part of expected cuts of 15,000 staffers if the merger goes through,he admitted, and R&D spending is likely to continue to slide from its current 5.4 percent of sales. However, Lampman is quick to add that in raw numbers the labs will grow in headcount and annual budget as HP embraces Compaq's research operations, which include remnants of the R&D groups of Digital Equipment Corp. and Tandem Computers.

Competitive spending

Lampman said HP's R&D spending is competitive with the industry norms in the diverse businesses the company is in — PCs, servers, printers and services. He cautioned that HP's R&D spending of 5.4 percent of sales should not be directly compared to IBM's 5.8 percent, because HP has a significant consumer business that IBM does not. IBM also spends significant sums on its semiconductor business, he said.

HP's chief executive Carly Fiorina spends a half day each month in briefings at the Palo Alto labs on the status of projects, pushing for them to set ambitious goals, said Lampman.

"When Carly was first announced as chief executive there was a significant amount of anxiety because she was the first CEO from outside HP and had a sales and marketing background," he said. "But it turned out the situation has been the exact opposite" of what some expected. "She has been a strong supporter of the labs and its role as a technology driver."

Researchers here were generally positive about the proposed merger with Compaq. "I think it's a good thing. Clearly there's a lot of consolidation going on in the computer industry. This is what we need to do to survive," said R. Stanley Williams, who directs HP's Quantum Science Research in molecular computing.

"I like the idea of becoming the second-largest information technology company in the world, with all the entailments that includes for innovation, reach and research," said Bernardo Huberman, an HP Fellow in information dynamics.