The Other One

Craig Barratt: Not to be confused with Craig Barrett, CEO of Intel. After a decade in the cellular industry, this one became VP of technology at Wi-Fi chipmaker Atheros last year. Recently named CEO of same, putting him in direct competition with... guess who?

I've never met the other Craig Barrett, but our paths have almost crossed several times. We were both at Stanford University at different times. We co-authored textbooks — mine in electrical engineering, his in material science — through the same publisher but at different times.

From time to time I get nice seats in restaurants, but other than that there hasn't been any significant confusion. One Craig Barrett thinks the Centrino chip is a great innovation, and the other Craig Barratt thinks the wireless part of Centrino is a step backward.

Intel is increasing public awareness of Wi-Fi, and we think that's great. But what they're doing today with Wi-Fi is very much old-generation technology. They have an 11b-only product, which doesn't have good power consumption and is not future-proof.

Atheros has a notion of a universal Wi-Fi client, a card you can plug into your laptop that connects to wireless LANs anywhere in the world, whether it's 11a, 11b or 11g. Moreover, our silicon also goes into the access points, whereas Intel just focuses on PC laptops. By controlling both ends of the link, we can provide our customers with better features: We have a turbo mode supported in 11a, and soon in g, that doubles the throughput.

Before Atheros, I was at ArrayComm for about ten years. My experience at ArrayComm tells me that the wireless carrier business is very slow-moving, with a very long product life cycle and technology deployment cycle. In contrast, wireless LANs follow the computer networking model, which has short product life cycles and rapid market adoption. Over the next couple of years, we'll see a cultural collision between these worlds. Carriers that learn to adapt will benefit from the synergy between the two, and carriers that move too slow will be at a significant disadvantage.

By the time I arrived at Stanford in the mid-1980s, the other Craig Barrett had left long before and was already at Intel. Now that I've been promoted to CEO of Atheros, perhaps there will be a little more confusion.