Semiconductor science can be confusing. The textbooks we studied at university tell us one thing and the real world often tells us another.

We were told that the emission wavelength of a laser depends on the material system on which it is based. Not so for the quantum-cascade laser (see p32), whose emission wavelength is dependent on its structure and not the material used.

Theory tells us that semiconductor lasers based on quantum dots should have remarkable temperature-insensitivity. But for many years industry was unable to turn this theory into practice. Manufacturing the quantum dots has been a challenge, but finally, 20 years after the prediction was first made, products are coming onto the market (see p30).

In the area of VCSELs, many did not believe that long-wavelength versions could find applications in the telecoms market; with today's shift in demand away from long-haul towards local area networks, however, these devices look set to take off (see p27).

Another area that looks promising for the next few years is tapered laser diodes. Textbooks tell us that ridge-waveguide-laser diodes give us good beam quality, whereas broad-area diodes give us high power. Now a German start-up company has shown that you can have both at the same time by using a tapered laser diode that provides high beam-quality and high power (p24).

The race for a silicon laser is also a classic example of the textbooks telling us it is not possible and researchers defying the odds. Groups all over the world are coming up with ingenious tricks to try and get silicon to lase. But Larry Coldren, a professor at the University of California, Santa Barbara in the US still feels that a pure silicon laser is highly unlikely to have the power efficiency required for practical applications (see p38). But who knows?