TABLE OF CONTENTS One of the truly huge risks in writing this kind of a book is in attempting to predict the future of the technology. I m going to try, though, on the assumption that several years from now the predictions will either appear prescient, or at least very interesting, if perhaps not both.
There are several important trends driving the current development of Fibre Channel and similar networking technologies, including the following.
The exponential rate of improvement in all technologies will continue. As applied to silicon chip area or processor performance, this is commonly referred to as Moore s Law, which predicts a performance doubling time of 18-24 months, but the same phenomenon can be observed in many other technologies as well, with comparable rates of improvement.
There are different rates of exponential improvement for different technologies. Recently, disk drive density has been improving even faster than processor performance, and Gilder s law states that the doubling time for performance improvement in long distance networking technology is even faster, between 9 and 6 months (or even 3 months, in some areas such as Wavelength Division Multiplexing technology at some time). Sometimes these different rates can cause revolutionary changes in how different technologies are used.
Further, there s sometimes even the observation of faster-than-exponential improvement, particularly in complex systems, as interrelated technological developments reinforce each other. An example of this is in Internet data caching technologies, where the usable bandwidth of WANs and servers can be tremendously increased, in effect, by caching...
