Business & Tech

hiawatha bray | tech lab

Google learns that cool isn’t cheap

In an undated handout photo, a technician installs optic fiber in Kansas City as part of Google Fiber services. Optical fiber, most often called just “fiber,” allows upload and download speeds about 100 times as fast as what is typically offered in the United States, but only a few areas in the U.S. have the option of fiber. (Google via The New York Times) -- NO SALES; FOR EDITORIAL USE ONLY WITH STORY SLUGGED CIR FIBER CONNECTIVITY BY KATE MURPHY. ALL OTHER USE PROHIBITED. --
Google via The New York Times/File
Google is dialing back its efforts to wire American cities with high-speed Internet.

I wouldn’t care to match IQ points against the people at Alphabet Inc., parent company of Google. Still, Google’s second thoughts about building superfast fiber-optic Internet service left me wondering what Google’s cofounders Sergey Brin and Larry Page were thinking when they launched the program six years ago.

Then it came to me: They launched Google Fiber mainly because it was cool. There’s a lot of that going on at Google, a company with enough billions in the bank to take the occasional flyer — four-legged robots, for instance, or educated thermostats.

But lately the company’s been backing away from some of its high-stakes investments, in a manner that suggests it’s starting to balk at the high price of cool tech. Especially now that we’re a few years away from a new wireless technology that will be just as fast as running fiber to our front doors, but a whole lot cheaper.

Advertisement

I love the idea behind Google Fiber. By seeding cities such as Austin, Texas, and Provo, Utah, with networks capable of handling a billion bytes per second, Google could give the entire broadband industry a jolt of healthy fear, and a little competitive adrenaline. The result: faster broadband for everybody.

Get Today's Headlines in your inbox:
The day's top stories delivered every morning.
Thank you for signing up! Sign up for more newsletters here

Yet now, with just half a dozen cities hooked up, The Wall Street Journal reports that Google is applying the brakes out of concern that the process costs way too much money. But how could the company not have seen this coming? Consider that it cost Verizon Communications Inc. $23 billion to string its Fios fiber optic network to about 18 million homes over the past decade. In Boston Verizon is vowing to replace virtually all of the city’s copper phone lines over the next six years at a cost of at least $300 million.

Verizon, unlike Google, is a telecom company born and bred. It already has thousands of miles of wire in the ground and it co-owns nearly every telephone pole in Boston. Still, Verizon will need hundreds of millions and most of a decade to get the job done here. Even for people who know what they’re doing, wiring a city is very expensive and very difficult. Apparently, this was news to Google.

But the company is not giving up altogether. Instead, Google plans to fall back on advanced wireless data systems to build the “last mile” of its future networks. Already a few companies offer such services. They attach a small antenna to your home or business and use this to swap data at broadband speeds.

Allston-based netBlazr offers this kind of service. In June, Google bought a San Francisco company called Webpass that delivers wireless Internet at up to 1 gigabit per second to office buildings and apartment blocks in five major US markets, including Boston. Taking Webpass nationwide is one way to get Google’s broadband ambitions back on track.

Advertisement

But there’s even better technology to come. A new generation of wireless services called 5G is capable of delivering fiber-like speeds to smartphones. Our 4G phones will generally give us 10 megabits of data per second, if we’re lucky. A 5G service could be 100 times faster — 1 gigabit through a wireless phone, the same top speed offered by Google Fiber. Indeed, during one recent test, a Verizon 5G system topped out at 3.7 gigabits.

This might sound like fantasy, but the Federal Communications Commission takes it very seriously. Last month, the agency freed up a huge chunk of underutilized radio frequencies to be used to deliver 5G services.

Still, commercial 5G networks are three to five years away, partly because they require a lot of new fiber. Today’s cellular networks already use fiber to handle our voices and 4G traffic. But existing fiber connections would be overwhelmed by the data from thousands of 5G phones.

In addition, 5G uses very high radio frequencies that have limited range and are lousy at penetrating solid objects — walls, for instance. Verizon and other cellular carriers hope to solve the problem by setting up hundreds or thousands of mini-cells, mounted on lampposts and utility poles all over town. Each of these mini-cells must be plugged into high-capacity fiber. That means putting more fiber into the ground. Verizon refused for years to bring Fios to Boston, but the unavoidable 5G network upgrade finally tipped the scales.

And 5G has gotten Google’s attention as well, showing the company a way to fulfill its broadband ambitions without having to take a backhoe to every street in America. It doesn’t get much cooler than that.

Hiawatha Bray can be reached at hiawatha.bray@globe.com. Follow him on Twitter @GlobeTechLab.