The Un-Trends in Photonics Markets

In the last post I ruminated on the best market trends of the decade in photonics. This time we examine that trends that weren't: the Un-Trends.

The decline of optical storage. Remember laser disks that were as big as vinyl records? Remember when CDs were displacing magnetic tapes? Remember CDs? The business of optical storage has been hit by a triple whammy: falling sales as downloads increase and falling prices due to commoditization. The iPad is the next big thing and--surprise!--there's no DVD player there. Not now, not ever. The future for optical storage is now in mass storage. And there's a chance that lasers will be needed to take magnetic storage a little further. Stay tuned.

The long winter of telecom components. The telecom components business never really recovered from the boom of the late 90s. Or shall we say it's back to the business it always was. Components. There are some successes, and stock prices are back to "normal," but overall it's nothing to brag about. Companies struggled through the decade to fill their fabs, move production to China, and just stay open. It's better now, but somehow it feels like there wasn't closure.

The elusive photonic integrated circuit. Twenty years ago it was called the OEIC, the optoelectronic integrated circuit. That was Bell Labs. In the 90s, Japanese companies pushed PLCs, photonic lightwave circuits. Now there are photonic integrated circuits (PICs) and the likes of Infinera and Luxtera. And of course there's the mother-of-all-quests: Intel's search for the silicon laser. (A 2006 article asks: lasers integrated into CMOS by 2010?) It's all nice work, and we're happy for Infinera--it's done a remarkable job. But the classical idea of the uber-circuit that will integrate smoothly with silicon--it's soooo 20th century. Get over it. What works is very piecemeal: some hybrid pick-and-place here, monolithic integration of a modulator there, and even then the economics can be questionable. The problem is that these approaches work best when you have high volumes, but high volume products are already commoditized in Asian factories. The real successes are much less dramatic. Think optical mouse, not 100G.

Optical computing is dead, long live optics in computing! I mean here the type of optical computing where the processor is all-optical. I worked on a project about this in the 1990s, with Japan, and there's still a little funding in it. The closest thing to it nowadays may be the all-optical telecom switch. (The name "switch" doesn't do it justice. They are pretty complex.) The optical processor is a nice idea ("it travels at the speed of light!") but it turns out that electronics is really really good, and really really cheap. Oh, and it's way easy to program. Now if you are talking about "optics in computing", that's another thing. There are optics everywhere inside a computer: the display, the mouse, the camera, the DVD player, maybe even a fiber cable someday (one can hope).

The death of CRTs, photographic film, and fax machines. The triumph of flat displays means the death of CRTs. What a great technology. Tubes are still preferred in various niches in electronics (take apart your microwave oven if you don't believe me). But I'm glad to see them go. Ever tried to carry a big CRT? (You can still buy one. Check Amazon.) Photographic film is still around too, although the last Kodachrome processor closed after Kodak stopped supporting the chemicals needed to develop it (see photos from the last roll here). It's hard to miss film, especially in the dentist's office or the hospital x-ray lab. And fax machines will still be around for faxing legal and medical documents, and for receiving wacky advertisements (does that ever sell anything?).

There's more, but isn't 5 enough?