Intel's CTO sounds a little bitter
had its big Research
Day yesterday at the Computer History Museum. I expected CTO Justin Rattner
to set the table by highlighting some of the futuristic developments in the
nearby exhibit hall.
Eventually he got around to some of that, but first, it seemed to me, Rattner
wanted to get a few things off his chest. He talked about the impact research
is having at Intel (NASDAQ: INTC), "not
only on the thinking and thought process, but the products and technology
forming the future of the company."
Then he noted the recent
introduction of Atom, Intel's highly power-efficient processor for mobile
devices. "I discussed with (CEO) Paul Otellini a few months ago that
we'd been working on low-power processors for quite a long time." He
noted at Atom's
big rollout "there was no mention of the very smart people who worked
for years" on research related to its development.
"There were at least three
attempts to sell the company on a low-power IA (Intel Architecture)," he
recalled, starting in 1999 when a group of researchers undertook "significant
work" to come out with much lower power versions of the Pentium line to as
little as a few watts. "It was reviewed by the senior staff but didn't
make the cut," Rattner said.
Work on the Atom began in 2004, and Rattner
said he's gratified it's been so well received -- finally.
"We believed in the idea
wholeheartedly," he said. "It's a clear example of a long-term,
persistent research effort ultimately having a big payoff. We believe Atom is
the fastest CPU in the sub 3-watt space."
He then traced similar long-haul efforts to when vPro and work on WiMax was
started -- 2002 and 1999 respectively. He said Intel had a vision for
fixed and portable mobile connectivity superior to cellular back in 1999. WiMax
"didn't just fall out of the sky" he said.
Giving new tech a chance
Later during a Q&A session, Rattner said in year's past
some technology would end up on a shelf for years "because no one on the
product side took a look at it." He said in recent years Intel's changed the
way research presents its ideas, now interacting more with business product
managers to see what's viable. "The hit rate has gone up
dramatically," he said.
So is Rattner bitter that some of the research efforts took so long to see the
light of day? I didn't get a chance to talk to him afterward, but at a
minimum, it sounded to me like he wanted to be sure the research crew at Intel
got some overdue love for its accomplishments.
Nathan Brookwood, Research Fellow at
Insight64, said Intel, like most other large tech firms, works on projects that
will never see the light of day. "It's useful for a company the size
of Intel to have programs going on because you never know when they'll need
He noted that Intel didn't appear to make a big push into lower-power chips
until after Transmeta burst on the scene. "Nobody at Intel was
interested in low power in 1999," he said.
Except a group of