H-P to unveil big revamp in its famed labs

Commentary: H-P walks fine line between basic research and market needs

By Therese Poletti

No technology company wants to end up with great research that it fails to commercialize.

Silicon Valley is too familiar with the failure of the research lab previously known as Xerox PARC to capitalize on its early innovations for the personal computer in the 1980s. Their work provided the seeds for the point-and-click user interface commercialized first by Apple Inc. AAPL, -4.62% and then Microsoft Corp. MSFT, -3.19%, and Xerox got only Apple shares.

“All the MBAs read about ‘Fumbling the Future’ at Xerox PARC,” said David Patterson, an inventor and professor in the computer science department at the University of California, Berkeley, referring to a book on Xerox PARC. Even though PARC went on to innovate in its core businesses, such as developing laser printing, its storied failure lives on in the Valley.

Amid this backdrop, Hewlett-Packard Co. will unveil a big push to ensure that its famed research group, H-P Labs, about a mile from PARC, contributes more to the printer and computer giant’s bottom line. H-P Labs, PARC, SRI and IBM’s IBM, -3.54% Almaden Research Center are the elders among the Valley’s research institutions that are now all confronting new ways to turn some of their research into commercially viable projects at a faster pace.

This Thursday, at an event hosted by H-P HPQ, -5.92% Chief Executive Mark Hurd, executives will sketch out new directions for H-P Labs.

This revamp is expected to be the biggest reorganization of H-P Labs in more than a decade, and details are expected to provide the key areas of focus for the future for one of the tech industry’s most prestigious research laboratories. The legacy of H-P Labs, founded in 1966, is also entwined with that of founders Bill Hewlett and Dave Packard, who were inventors themselves and whose offices are enshrined at H-P Labs’ local home, an airy modern building on Page Mill Road, a stone’s throw from headquarters in Palo Alto, Calif.

Thursday will also serve as a sort of debut event for Prith Banerjee, a former academic and entrepreneur who left life in the ivory tower to join the new H-P under the regime of Hurd’s cost-obsessed and results-oriented H-P. So far, one source said, Banerjee has been removing layers of management in the labs, and projects without results in sight are getting shelved.

In an era of cost-cutting and ruthless focus on the bottom line, it is perhaps miraculous that H-P still conducts basic research into areas such as quantum science, nanotechnology and information management.

“With the turmoil that they went through, it is a real credit that they kept the labs together,” said Paul Saffo, a veteran technology forecaster. “H-P Labs has always been one of the great corporate R&D labs.”

In recent years, some innovations out of H-P Labs include its major advancements in developing computer chip circuitry at the atomic scale, software to automate data centers and building what is believed to be the first utility computing center, where customers can get computing power based on their changing needs.

I can only guess what some of H-P’s big focuses will be, based on a chat I had with Banerjee last year. He wants to identify big themes, really big bets, and throw a lot of resources at those top areas. H-P will likely continue to focus on nanotechnology, reducing the power and increasing automation in computer data centers, and advances in digital printing, among others. Eventually, H-P will focus on about 30 projects, Banerjee said last year.

Banerjee, who appears to be well-liked among his H-P Labs colleagues and respected among academics, made it clear in a few interviews after he joined H-P that he plans to focus all research with a specific market in mind.

“I’d start with making sure that when a research project begins there’s a reason why – not just a scientific reason, but a market research, why this work should proceed,” Banerjee said last year. He is also expected to have staffers from the product divisions join the research teams within the labs, to move faster toward creating real products, when feasible.

H-P spokesman David Harrah declined to provide any specifics on what H-P Labs will announce, only to say that H-P will discuss the direction of the labs. The event originally coincided with a press conference at Apple in Cupertino, where the company will launch its roadmap for developers to help them create software applications for the popular iPhone. H-P originally had scheduled its event an hour before Apple’s and later changed it to the afternoon, to accommodate Hurd’s schedule, Harrah said.

I suspect they moved the time after they saw the competing event from Apple, but that’s just me.

Like many other tech companies faced with the dilemma of how to innovate for the future and keep generating revenues to fuel the bottom line and fund that research, H-P is also likely to adopt a model of more partnerships with universities, given Banerjee’s background at the University of Illinois, where he was the dean of the college of engineering.

Forming partnerships is a trend that many tech firms have embraced in recent years. Microsoft, for example, which formed its now-respected Microsoft Research in 1991, will open up a new research lab later this year in Cambridge, Mass., allowing for more collaboration with nearby universities such as MIT and Harvard. Intel Corp. INTC, -3.89% has formed small lab-lets with engineers on university campuses, such as Berkeley.

It is a positive trend that lets the tech business innovate at a faster pace while at the same time government spending on basic research continues to decline.

“There has to be a shorter time span between innovative ideas and that tech transfer,” said David Bader, Executive Director of High-Performance Computing at Georgia Institute of Technology in Atlanta, which works with IBM, Sony and Toshiba on the Cell chip. “It changes the traditional model of doing research and doing a slow handoff to industry and seeing it in a product in 10 to 20 years. Instead it’s down to two and half to three years.”

As it winnows out its projects, H-P also has to walk a fine line of focusing on research that bears fruit more quickly, and not scare away their star researchers who may be disappointed with the loss of their pie-in-the-sky projects.

I am not sure completely cutting far-out projects with no end in sight is such a good thing, because sometimes this is where hidden gems may lie, but hopefully Banerjee is harvesting the best out of H-P Labs.


David A. Bader
David A. Bader
Distinguished Professor and Director of the Institute for Data Science

David A. Bader is a Distinguished Professor in the Department of Computer Science at New Jersey Institute of Technology.