I suspect, with the reduction in size, Dr. Taylor's team has discovered a few things that at larger nodes did not occur. I recall the mention in a NR that achieving 100nm has shown to be more challenging than originally anticipated. After 20+years of theory and large scale development, the realities of small scale versions can become quite vexing. I believe that is why they have had to "find the path" to 100nm as each drop in size created both anticipated and unanticipated limitations/challenges. Hence the 20x performance figures rather that the much touted 100x improvement originally forecast. Once they have ironed out all of the wrinkles to get maximum performance and reliability out of POET, I believe, what will be presented will still be a significant leap in speed and other metrics from today's chips. It just may be less than the 100x originally projected. Still a great place to start for a new technology. Looking forward to what comes. D
