The Holy Grail of Optical Networking

The Holy Grail of optical networking is the development and deployment of passive optical circuits, but the technology may have an impact on personal computing as well. Most optical circuits today are active, meaning that electrons and photons are converted back and forth. The passive optical circuits are designs that directly manipulate photons as if they were electrons.

It comes as no surprise that IBM, whose microcircuits division has been pioneering all sorts of developments, is a potential developer of passive optical circuits. When IBM's key invention, a material known as siliconoxynitride (SiON), is used as a "waveguide layer," passive optical switching circuits can be developed. If any of this can be standardized as the CMOS process was, it may become the basis for desktop computing of the future.

IBM has teamed up with Kymata, a fast-growing Scottish company specializing in optical technology. This company seems to have at its roots a technology transfer from British Telecom in the early 1990s. It is a DWDM (dense wavelength division multiplexing) specialist with engineering and sales offices throughout the world. The company is privately held, and it boasts an elite list of investors, who have provided $160 million in funding so far. Kymata, showing signs of being a mini Cisco, has been buying up optical-circuit specialists, and it is one of a number of U.K. and European efforts to leapfrog American technologies at the highest levelUnfortunately for the Euros, the highest-level technologies always seem to boil down to ultraminiaturization. Hello, Switzerland! Even Kymata is involved with MEMS (microelectromechanical systems), for example. The company is part of a Pan-European consortium started by the EU called Europractice, which defines itself as "a European Union initiative to improve the competitiveness of European industry by the adoption of advanced electronics technologies." The conferences established by Europractice put an emphasis on sensors, transducers, MEMS, and nanotechnology. Europractice seems to be focused on collaboration between companies. Meanwhile, the European Union managed to quash the deal between GE and Honeywell. The way I see it, there will be trouble a-brewin' over all this, one of these days.

Another Kind of Optical Technology Dept.: On the other side of the globe, Fujitsu is bucking a trend by almost single-handedly maintaining a market for high-performance magneto-optical storage systems. (Remember MO?) The company just introduced a 2.3GB version of its 3.5-inch optical drive. It is fully backward-compatible with all the older 3.5-inch drives, including the once-jazzy 128MB drive. This is a technology I revisit every few years as I wait for its demise. What keeps it alive is the incredibly rugged nature of the storage media.

According to Fujitsu, the drives are still needed for reliable removable read/write systems such as those used for medical imaging. Also, the R/W performance can now be used for video editing and other high-end purposes. The newest technology employed by these drives is MSR (magnetic super resolution), which allows a bit to be written by a laser that is actually smaller than the beam itself. This was thought to be impossible a few years ago.

Genuinely Interesting Software Dept.: Although few of us will be using the magneto-optical drive, almost everyone has a CD burner. I was pleasantly surprised to stumble across the highly regarded NTI CD-Maker, from Newtech Infosystems. The new Professional Edition is a superbly elegant system that simplifies the process of CD creation and duplication to an extreme. This software is close to perfection. It also lets you make VCDs (video compact disks) that can be played in DVD players. Hot stuff. The Professional Edition sells for $70 boxed, or $50 downloadable from the Web site (www.ntius.com).

Dot-Com Doldrums Dept.: As you know, this column has not followed the faddish dot-com scene to any extent, since it always seems like an anomaly. But I couldn't help chuckling over the demise of Webvan and its ability to burn through $830 million without making a profit. The CEO then said that if the company had more money and more time it could have worked. How much money do you need? The stock finally collapsed to a few pennies. When the stock was selling for $6, almost all the analysts had it listed as a strong buy.

Since the demise of the milkman 20 or more years ago, I couldn't understand how anyone could think this idea—a glorified milkman without a fixed route—could ever work. And although Webvan's quality was good, its prices were high. This was point number two: What are the reference points for modern retail selling, groceries or otherwise? Megastores with low margins. People aren't packed into Costco and Wal-Mart because they are looking for high quality at high prices. All I can conclude is that somebody at some group-think meeting must have given one helluva sales pitch to suck out $830 million in investment money against all logic