Trouble in Fibertown

Orem, Utah is a storied city in the annals of networking because it was the birthplace of Novell, the company that dominated computer networking in the ‘80s and ‘90s. Novell’s Netware products helped transform corporate computing from a centralized, mainframe model to decentralized, PC-oriented model.

Ironically, the company’s assets were recently acquired by the mainframe software company Micro Focus and then merged with the Hewlett-Packard Enterprise Division’s software unit. The demise of Novell was more fascinating than its rise in some ways.

Searching for UTOPIA

In 2002, Orem and group of neighboring cities began discussions that led to the creation of the Utah Telecommunication Open Infrastructure Agency, better known as UTOPIA. UTOPIA issued bonds backed by sales tax revenues with 11 of the 16 original cities participating.

The project broke ground in 2008, and remains about half done. It’s an understatement to say the project has been financially troubled from the start; here’s an analysis from the New York Law School’s Advanced Communications Law and Policy institute (page 75):

The cost of UTOPIA has been very high: factoring in debt service and other payments, the total cost of the network approaches $500 million. Of this, $185 million stems from long-term bond debt; the cost of the infrastructure itself was $110 million. Construction delays and lack of consumer interest required the network to use a significant amount of its bond proceeds to service its debt ($48 million) and make up for operating deficiencies ($27 million).

Having failed to sell the project to the Macquarie company of Australia in 2014, backers now propose to go back to the Orem voters for permission to take on another $40M in debt (including interest) to complete the project within their city at an accelerated rate.

Poster Child for Government Excess

Orem is served by CenturyLink and Comcast with the same options these companies offer in most places: up to 80 Mbps VDSL for the former and 250 Mbps cable modem for the latter. Consequently, the proposal to take on additional debt for UTOPIA is controversial.

Since 2008, UTOPIA has become the poster child for the poorly conceived and poorly executed government owned, taxpayer supported broadband project. It’s now been studied more closely than probably any such project except Australia’s National Broadband Network, a project with a similarly checkered history. The lesson is that building the network of tomorrow today means you have to wait until tomorrow to find customers, and even then they’ll only show up if your crystal ball gazing was correct.

And in these two cases – UTOPIA and the Aussie NBN – it wasn’t. UTOPIA was born in 2002, long before mobile devices had any use for broadband speeds, when TV was delivered over a customized network, and the TiVo was known only to a small following of techno-geeks. The rise of 2G, 3G, and 4G cellular changed all that, and 5G will change it even more.

We Were All so Naïve Once

There was a lot of enthusiasm for FTTH in 2002, most concretely at Verizon where the FiOS network was born. Like UTOPIA, Verizon believed their network was going to be so great that it would gain take rates of 40% right out of the gate and Wall Street would gush with optimism and push Verizon stock to astronomical levels. The company was able to roll out to the most promising markets quickly and economically because it already had poles and conduits in place. It quickly mastered the art of connecting homes to poles and aggressively marketed the product.

But the take rate didn’t materialize for a decade after the initial build, and the product ended up being much more expensive than imagined because the public still sees Internet as part of the double-, triple- or quad-play bundle that also includes TV. As a small player in the TV market, Verizon has never been able to negotiate the kinds of deals that larger players can make.

Inside the home, most of the devices connected to the FiOS network are wireless. So FTTH in practice is actually fiber to the router. Wi-Fi supports up to 1.3 Gbps now, more than enough capacity for the dozens of consumer devices attached to the typical home network now. Wire can scale to hundreds of Gbps, but speeds in that range don’t really matter inside the home even if they’re valuable to network-intensive businesses.

Orem’s Unique History

Orem is an important landmark to those of us who cut our teeth in the networking business in the 80s and 90s because it was the birthplace of Novell. When we designed Wi-Fi in the early ‘90s, Novell departmental file and print servers were the target application. Sadly, Novell failed to appreciate the importance of new (to the public) technologies such as the Internet, Windows 95, and mass-market distribution channels.

It seems that UTOPIA has suffered a similar fate in refusing to appreciate the importance of mobile devices, wireless IoT devices, and the needs of smart electric and transportation grids. Instead of pretending that nothing has changed since 2002, it may well be wise to take stock of today’s realities and plot a new course that builds on everything we’ve learned.

Policy makers in Orem are still engaging in the turn of the century debates about top-end speeds, and some of the ISPs who offer service over UTOPIA offer 10Gbps connections.  Like all wired connections, these services don’t directly support mobility, so their value is limited. But some lawmakers are reluctant to admit that FTTH will be all but completely supplanted by 5G networks a few hundred feet from the home with fiber backhaul.

Saving the Investment

Verizon has made FiOS into a nice business by making it a complement to its mobile business. Fiber installed for FiOS works quite well as backhaul, so 5G is going to be an easy upgrade in FiOS territory.

While most of the US is not FiOS territory, the learning, the parts, and the volumes will help everywhere. Verizon is set to purchase some 37 million miles of fiber shortly, and most of it is for 5G.

UTOPIA wasn’t designed to complement mobile networks, so it’s unclear how much of its existing infrastructure will be useful. UTOPIA is also shackled by its business model, one that has given rise to more than a dozen boutique ISPs that can’t offer competitive TV plans. UTOPIA homes probably get TV from satellites and skinny bundles.

Accepting the Reality of 5G

One of the options for UTOPIA would involve exploring the options for using its infrastructure for 5G residential service. This is probably few years off, but Verizon and others are running trials for this sort of service today. It’s not too early to begin exploration.

Another option would be to forget about residential service altogether and focus on mobile backhaul. It’s likely that much of the existing fiber plant is in good locations for small cell backhaul. 5G also opens up a lot of “smart grid” and “smart city” applications.

The alternative for UTOPIA is to push forward with $40M in bond debt to wire the other half of the city. Given the low take rate for UTOPIA – less than 28% at present – pushing forward with the status quo doesn’t seem productive.

Users quite likely believe the network is dandy, which it probably is, but that the small ISPs are not what they need to be to run a double-, triple-, or quad-play business in 2018. No amount of taxpayer debt is going to change the fact that networking is an economy-of-scale business.

When faced with the need to either stagnate or grow, Novell chose the status quo path. Let’s hope Orem doesn’t repeat the error with UTOPIA. It might have been a great idea in 2002, but the visions many of us had of networking in those days were blind to the progress that was possible for wireless. That was a serious miscalculation.