Tuesday, September 5, 2017

Deep Fiber Yes, FTTH Maybe No

It is safe to say that "fiber deep" access networks are going to be essential in the U.S. access market. What is not clear is the actual deployment pattern, the required magnitude and the timing.

By some estimates, as much as $130 billion to $150 billion in additional optical fiber deployment in the access and distribution networks to support U.S. 5G and allow telcos to compete with cable TV operators, over the next five to seven years.

Some of us doubt that will wind up being the case, at least not over the five to seven year timeframe. There are several reasons. First, and most compellingly, there are rival strategic claims on available capital.

Service providers also must make strategic acquisitions to gain scale and also move up the stack into the applications and platform spaces. Tier one providers that have made big acquisitions have to pay down debt from making those deals.

Stranded assets are another issue. In the facilities-based U.S. market, any tier-one fixed network service provider has to expect to strand at least 50 percent of newly-deployed access assets. No rational executive is going to invest billions, or scores of billions, knowing that half those assets will not generate revenue.

Consider “where” new optical infrastructure is called for, according to Deloitte.

According to Deloitte, $15 billion to $20 billion is required to support densification of the mobile networks (5G small cells).

Some $35 billion to $40 billion is needed to connect rural residents. About $60 billion to $100 billion is needed to bring fiber-to-home networks or other gigabit internet access speeds to residences.

Some might argue that the capital to support mobile network densification ($15 billion to $20 billion) is essential and affordable.

What is questionable, given the opportunity to use 5G-based fixed wireless, is the $60 billion to $100 billion in fiber to home, where half the investment will be stranded, and not generate revenue. This bucket of spending is questionable.

The $35 billion to $40 billion to support rural internet access likely will prove amenable to other solutions than fiber to the home, namely wireless in several forms, or mobile networks, using fiber trunking but not fiber to the home.

The point is that huge amounts of capital will be needed to make strategic acquisitions for scale; others for scope (moving up the stack); to pay down debt and then to invest in access networks and spectrum.

Under such conditions, no rational executive is going to strand so much capital in the access network. For that reason, one might argue that the $130 billion to $150 billion investment over five to seven years is too high. Faster speeds (hundreds of megabits to gigabits per second) can be supplied in other ways that are far more capital efficient.

No comments:

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...