Intel offers a glimpse at its technologies of the future
I spent most of Thursday hanging out at Intel's Berkeley, Calif.-based research lab where the company held an open house to showcase what it's working on. Berkeley is one of three "lablets" that Intel operates in association with two other universities. (One is in Seattle, where the company collaborates with the University of Washington, and in Pittsburgh, where Intel works with researchers at Carnegie-Mellon.)
Some cool projects were on display, though like most cutting-edge work, the odds are that most of this stuff will wind up on the cutting-room floor. No matter. It still offers a great look at what Intel's rocket scientists think may some day turn into real products.
I'll get to the specifics in a moment. But in conversations with several Intel executives, one theme came through repeatedly: they envision a tech landscape increasingly dotted by products which incorporate sensors and situational awareness, or inference capability.
"A lot of places are crying out for this in consumer applications and small mobile devices that you keep on yourself all the time," said Andrew Chien, a research director at the lab. "There's a lot excitement in living room. These are starting to burst out in gaming systems...with the kinds of sensors and gaming environments which will tap these sorts of enhanced sensors."
With that as brief prologue, here's a thumbnail sketch of the highlights.
Common Sense In the last year, Intel began an experiment to measure air pollution in Ghana's capital of Accra. Sensors fastened onto taxicabs fed back data on the differences in pollution levels across the city. The experiment was so promising that by the spring, San Francisco street sweepers will be similarly equipped with these sensors, essentially turning their vehicles into mobile pollution data collectors. As the sweepers go about their jobs, the meters will register street-by-street data on pollution levels. So now, instead of guessing about the environmental impacts of a particular urban policy, municipal agencies will have hard data to work with. Yes, they still may make dopey decisions, but they won't be able to claim they didn't have scientifically valid data.
(Credit: Intel)Mash Maker Intel's developed a software tool running inside the Web browser that will help non-experts assemble mashups on the fly as they surf the Internet. A year ago, Intel first showed the technology to the public, but it was server-based. Using a PC, users can write and customize widgets in the time it takes to copy and paste data from one Web page to another. When I timed the demo, it took about a minute from start to finish. Eric Brewer, who runs the Berkeley Lab, called this building "mashups for the masses." OK, he's obviously hyping for effect. But I think a lot of non-technical users will find this to be a useful tool to combine database results. Best of all, you won't need to wait until your turn in the corporate queue for a developer to fill out your request.
Proteus This is a software security project that started out by comparing how viruses behaved in the human and computer realms. In biology, viruses don't transfer from species to species because of differences in DNA and naturally occurring diversity. The question put to Intel researchers was whether there was a relevant analogy to apply in the struggle to stem the propagation of computer viruses.
So for the last one-and-a-half years, Intel has concentrated on looking for ways to stymie hackers. Researchers think they've made a good start. The idea is to make each computing platform in an IT setup sufficiently different so that viruses can't hopscotch from one category of computer to the next. Intel's now working on algorithms which would automatically configure those differences within an IT department. No doubt, it will have to be as close to automatic as possible to succeed. Budgets are already tight and that's not likely to change as the year unfolds. The people I spoke with were optimistic about releasing this to the public soon, perhaps within the year. Still unclear is whether it winds up in firmware, or chip hardware.

Hardware for Intel's rural connectivity platform
(Credit: Charles Cooper)Intel's rural connectivity platform We wrote about this a year ago. Now it's almost ready for roll-out sometime this fall. It's a low-cost way of providing roughly 10 megabits-per-second connectivity to remote areas. The RCP needs a clear line of sight to work, but when set up without obstructions, Intel says the wireless long-distance nodes can connect every 60 miles. (I was told that one test offered up 4 megabit speeds up to 100 miles away.) Intel says it tweaked the Media Access Control protocol with a TDMA modification. The hardware is comprised of a small box which includes a single computer board, a lower power processor, a compact flash card for storage, and mini PCI wireless cards.
Charles is an executive editor with CNET News. He has covered technology and business for more than 25 years. A graduate of Queens College and Columbia University, Cooper began his career in journalism at the Associated Press before moving to technology coverage. Before joining CNET News, he worked at Computer & Software News, Computer Shopper, PC Week, and ZDNet. He received the Excellence in Journalism award from the Northern California branch of the Society for Professional Journalists for column writing. In addition to his blogging and podcast appearances, he is a co-host of the CNET News Daily Debrief. E-mail Charlie.






the more money they spend on software dev, the more they waste on things that smaller companies and groups of programmers will end up doing better and cheaper, if not free.
make the cpu smaller and faster, pour money into r&d for optical circuits...stop wasting your money on software and dev gadgets that have little to do with engineering chips...let the gadget people do their thing, you do yours.
if intel doesn't focus, it'll get blindsided by competition that it doesn't realize is even there, until after the fact...just as amd did to it with the athlon.
the more money they spend on software dev, the more they waste on things that smaller companies and groups of programmers will end up doing better and cheaper, if not free.
make the cpu smaller and faster, pour money into r&d for optical circuits...stop wasting your money on software and dev gadgets that have little to do with engineering chips...let the gadget people do their thing, you do yours.
if intel doesn't focus, it'll get blindsided by competition that it doesn't realize is even there, until after the fact...just as amd did to it with the athlon.
There are several of the new Intel CPUs I'd like to purchase such as the q9550 - but on all the websites I've been been seeing in stock soon, preorder now or shipping in a week for over two months now.
I contacted Intel to find out what is going on and officially the CPUs were released in January - but they wouldn't comment on why no one has them in stock.