I was in the White Hart pub in Godstone a couple of days ago having a beer with Kevin Kalkhoven. He was over in the U.K. closing a couple of startup deals and had popped in for a couple of pints of Boddingtons.

Kevin's been having a lot of fun since he left JDS Uniphase Inc. (Nasdaq: JDSU; Toronto: JDU) last May. He’s taken up motor racing again. He’s buying a bigger jet, and he plans to fly it around the world later this year. He’s also into scuba diving and skiing (see Kevin Kalkhoven ).

But that’s just for starters. When Kevin hasn’t been racing, flying, diving, or skiing, he’s been putting together a grand plan to repeat the success he scored at JDSU. In fact, he’s hoping to do even better this time around.

Kalkhoven’s success at JDSU all started when he was heading up Uniphase, a company making gas lasers. At the time, back in 1995, he recognized the need for an independent source of optical components, so that folk like Ciena Corp. (Nasdaq: CIEN), Cisco Systems Inc. (Nasdaq: CSCO), and Sycamore Networks Inc. (Nasdaq: SCMR) wouldn’t have to buy their components from competitors like Lucent Technologies Inc. (NYSE: LU) and Nortel Networks Corp. (NYSE/Toronto: NT).

To cut a long story short, Kevin and Dan Pettit, his CFO, formulated a five-year plan to transform Uniphase into the world's first independent merchant supplier of optical components. This involved identifying all the different types of widgets that would be needed in next-generation telecom equipment and then going around the world, acquiring the capacity to make them.

As we all now know, the plan turned out to be wildly successful. It propelled JDSU to multibillion-dollar sales and gave investors a lovely ride. It also forced companies like Alcatel SA (NYSE: ALA; Paris: CGEP:PA) and Lucent to spin off their component subsidiaries, leading to restructuring of the whole telecom equipment industry.

Kalkhoven now reckons he can repeat the process on an even wider scale. He says he’s got a clear idea of where telecom will be in five years time, so he and his pals –- Dan Pettit and Jack Levin -- are setting up companies and investing in startups that can turn this vision into reality. This is being done with their own money, via a company they’ve set up called KPL Ventures (no Website). This arrangement allows the trio to take bigger risks than conventional VC outfits that have to answer to their investors.

Kalkhoven's big thing is access networks. He predicts that within five years broadband cellular technology will have blown away all the other ways of providing broadband connections. He sees two-way video becoming the big traffic generator and says asymmetrical technologies such as DSL (digital subscriber line) simply won’t cut it. He also figures folk will have come out with some very low-cost ways of providing fiber connections to businesses. Access networks represent a huge potential market -– much, much larger than the potential market for carrier backbones, he adds.

In general, KPL is putting its money into startups developing components that'll make this happen. Its investments include Blaze Network Products Inc., which is developing coarse wavelength-division multiplexing widgets for low-cost access networks (and, incidentally, is already selling them to Cisco); and LGC Wireless, which is developing broadband wireless technology. KPL’s also invested in WaveSplitter Technologies Inc., which makes AWGs (arrayed waveguide gratings). And, as already noted, Kalkhoven is helping found a couple of component startups in the U.K. He promises more details later this year.

KPL is also investing in Innovance Networks, which is out of character in that it’s developing systems rather than components. Kalkhoven says this gets him inside the brains of systems developers to figure out what components to focus on. It also gives his other startups somewhere to try out component developments.

Later on in our lunch-time tête-à-tête, Kevin dropped another little bombshell that had me choking on my banger. He thinks that a lot of people have got totally the wrong idea about the way the core of telecom networks will evolve. He sees everything becoming a lot more distributed so that there's no need for very high channel-count DWDM systems and monster optical switches. In fact, he thinks that optical switches aren't going to take off, full stop. He's putting his money into tunable components (he's an investor in Iolon Inc.), which can be used to make reconfigurable networks that don't use optical switches. (For an example of how this might work, see Researchers Unveil All-Optical Advances.)

As broadband access networks get deployed, Kalkhoven believes, folk will bypass expensive old-fashioned architecture backbones, and many traditional operators will go out of business. In the long run, however, broadband access networks will give the surviving backbone operators a second lease on life -- history repeating itself, he says. The PC killed off nearly all the mainframe vendors. Only IBM survived. Now, however, mainframes have found a new role in life, because PCs have fueled an information explosion. In just such a way, the broadband explosion will undermine the traditional telcos -- such as AT&T Corp. (NYSE: T) and Sprint Corp. (NYSE: FON) -- while opening out new vistas of opportunity. Those backbone operators that survive the revolution will thrive, providing the huge amounts of capacity required to interconnect a new breed of broadband service providers.

-- Peter Heywood, International Editor, Light Reading http://www.lightreading.com