Innovative research is still supported in the here, it's just less obvious, said Gordon Bell, during a panel session here Wednesday.

"We're funding as well as ever. We've just got all these people who are all coming to the funding feeding trough," he said.

Bell, principal researcher with Microsoft Corp., was one of several long-time Silicon Valley luminaries reunited for Wednesday's event, held at the Computer History Museum to honor the 40th anniversary of Bob Metcalfe's original Ethernet memo. He was appearing on a panel about the role of architectures in guiding innovation, but he took a second to float a contrarian view about the state of innovation.

The day was filled with big names from the early days of networking, including Metcalfe and many of his Xerox PARC compatriots, including Bill Spencer, former PARC head, and Radia Perlman, inventor of the spanning tree algorithm and now an executive vice president at Intel Corp..

The question is whether it's a problem that the institutions of basic research -- Xerox PARC, where Ethernet got its start, or Bell Labs -- have been diluted, usually victims of corporate belt-tightening. Conventional wisdom says this is a problem, but Bell disagrees.

"There's a constant set of great ideas and constant set of really dumb ideas and 'How-could-you-start-a-company-off-that,' not to knock on Facebook or anything," he said. "I just see a lot of great ideas ahead ... and we'll call a lot of them Ethernet."

Bell's was a minority opinion at the event, though, and probably isn't shared by many in the Valley.

To be sure, any scraps at the trough have gotten smaller. Paul Grams, Exploration Technology Directorate at NASA, talked about page after page of programs listed by the National Science Foundation, most of them for grants of $500,000 or less.

What's gone are the larger institutions where pure research was pursued for its own sake, in an environment almost completely unfettered (or so the legends say).

What made those institutions fruitful was the step of moving their ideas into the commercial sector. Ideas could be nurtured, iterated and smoothed in a place like PARC, then brought into a practical reality by a company like Bridge or 3Com, said Judy Estrin, CEO of JLabs and founder of Bridge Communications.

"We're missing this transitional piece," Estrin said. "We need to be looking at that whole cycle, or 20 years from now we're not going to have that innovation we had ... 40 or 50 years ago."

"We really have to go back to the discovery phase and have discovery convert itself to innovation before entrepreneurs can take that innovation and bring it to the mass market," said Yogen Dalal, another member of PARC's original Ethernet team and a partner emeritus with Mayfield.

Universities can't always complete that handoff, Stanford being a notable exception. And commercial industry, with its necessary short-term focus (think quarterly earnings reports) is increasingly skittish about letting research ideas simmer the way PARC did.

The panel didn't come up with any magic answers to the problem, which isn't surprising. It's a hard problem. But throughout the discussion, there was an optimism about the level of knowledge and technology development still percolating in the U.S.

China might do a better job at funding basic science, "but not [at] the whole cycle," Estrin said. "I still think we have an edge."

For more

• Joyeux Anniversaire Ethernet
• A Walk in the PARC
• PARC's Past & Future, With Bob Metcalfe

— Craig Matsumoto, Managing Editor, Light Reading