Components > Advanced Semiconductors Blog


Nationalizing 5G? And Rationalizing Compound Semiconductor Demand?

by Eric Higham | Jan 31, 2018

Hello and a belated Happy New Year to everyone. I have to admit that I have been conspicuously absent from publishing blogs, or really much of anything the past few months. I’ve been head down and chugging away on some very interesting projects, hoping to keep my ears and eyes open to new developments. Imagine my surprise when I uncovered a bit of politically tinged information on Sunday. Now, I try to avoid talking about politics as if it’s the plague or this new strain of flu that is running rampant in the US. Nothing good seems to come from politics these days; it pits friend against friend, brother against brother, lover against lover and yet, here I am, about to talk politics.

To set the stage; there I was on Sunday, two days before the first State of the Union Address from a first-term president to a distinctly divided country. There are several topics that seem unlikely to get bipartisan support and I stumble on one of them in my e-mail inbox. Of course, I’m talking about the leaked…not leaked…plan…non-plan to nationalize 5G! What else would catch my eye? I’ve waited until today to write about this and no, the timing has nothing to do with the State of the Union Address taking place last night, but rather to give the industry a chance to weigh in this development.

In case you haven’t seen the news, the online news site Axios claimed that they had gotten documents detailing two options under consideration by the administration to nationalize the 5G network in the US. The first option has the US government funding and building a single network without the aid of operators. The second option was to mediate some sort of cooperative agreement among all the operators to help speed network deployment.

As I expected, the response to this story was swift, served to stir up the already muddy waters and yes, political. Telecoms.com ran a story headlined Paranoid, egotistical and naïve…  and the story wasn’t any more flattering than the headline. Other media outlets have suggested that the plan has evolved and the second option now aims to set aside spectrum and give it to some “uninterested third party”, like Google (as if they are an uninterested party to anything!). Google would manage the development and the deployment and rent capacity to the operators. There was even an indication that the idea had evolved to the point where Google might provide some bandwidth to users to innovate. I also had a contact send me a link to a story that claims this effort envisions a unified defense network with excess capacity available to commercial operators, for a price.

The proponents of nationalization point to the increasing need for security, especially when so many of the applications and the technology will reside in the cloud. China seems to rise to the top of this discussion and the apprehension about cyber security makes this a valid concern. Another big concern is the time needed to deploy the 5G network and the implications for the “5G race”. Coupled to the speed of deployment is cost; how much network capex is required and can the operators develop a workable business model around 5G.

The opponents of the plan are pointing out the US government is seldom the most efficient contractor for anything. They question whether this is an appropriate use of (many) tax dollars and whether the federal government should be inserting itself into local issues like site location, zoning, etc. Lastly, they question whether operators and shareholders can forgo the promise of profits in the name of cooperation and national security.

Good points on both sides and this got me thinking. The US government certainly does not come to mind when efficiency, speed and the ability to get things done are important. But, one thing they do very well is spend money!

As a quick refresher, one of the first deployment cases for 5G will be the non-standalone (NSA) model, where the 5G network will leverage the existing 4G network. The graphic below shows the 4G coverage maps for the four major US carriers. From these coverage maps, it’s clear that if you’re in Alaska, you may not want to count on 5G any time soon. Sprint has more holes in their coverage than they have coverage and all the carriers have gaps where the terrain gets a bit rugged. Verizon’s webpage touts coverage of 98% of the US population and 2.4 million square miles of the US land mass. They claim this is 400,000 square miles more than their closest competitor covers.

4G Maps

This is all very interesting, but in Verizon’s case, their coverage is the result of 17 years of deployment and $110 billion dollars of capex. In addition, various sources list the total US landmass at ~3.8 million square miles. That leaves ~1.4 million square miles with no cellular service. The problem gets a bit thornier because Verizon has a great interactive coverage map on their website. I can zoom the map down to my street level and my 4G coverage could not be any more robust, according to Verizon. The problem is; without a 3G network extender, I don’t get ANY service at my house.

Let me revisit a blog avatar that I’ve used frequently. One of my favorite people in the world, my cousin April, lives in a tiny community on a beautiful lake in the Adirondack Mountains. Half the year she has snow and the other half she has mosquitoes, but that’s another story. The community gets no cellular service, but there are a couple of small businesses in the area that have convinced a telco to run DSL in. The central office is very close to her house, so April gets 20+ Mbps DSL internet service.

Now, imagine that we have achieved the 5G vision and everyone has virtual reality, everything is connected to the cloud and autonomous, everyone has fast enough internet to enable all the body telemetry and personal areas networking needed to improve longevity and quality of life, etc. Well, everyone except for April…and me at my house…and anyone traveling through the 1.4 million miles of this country that are not currently served by a network. What happens if I want to visit my cousin April and I lose connectivity along the way. I guess I better be paying attention so that I can take over driving my car and I had better not have a medical emergency, because who would know?

But, what does this have to do with the compound semiconductor market and the nationalization of 5G? In a survey of operators that I used in a presentation two years ago, 76% of operators said they had no plans to deploy a 5G network. The biggest reason for this was a lack of clarity on the business model. That may have changed, but it is unlikely that the business model ever works for April’s house, or my house, or the 1.4 million miles of land in the US without a network. Without that coverage, there is no ubiquitous 5G network.

This is where the “thought exercise” gets interesting. The best way to cover those areas that don’t currently make sense may be for the government to fund the efforts. This means more coverage and a faster deployment schedule, which is good for the compound semiconductor industry. It also means a single network, which will reduce the equipment quantity.


Many of the responses compare 5G nationalization to President Eisenhower’s interstate roadway effort started in the 1950s. The articles I’ve seen belittle the nationalization thought, correctly pointing out that an effort of this magnitude hasn’t been undertaken in 60 years. But, the need for superhighways that opened up commerce in areas that were hard to reach have given way to the need for information superhighways that can similarly open up commerce in areas that are currently underserved. Perhaps it’s time to think about how we can enable this effort, rather than why we can’t.

Long blog, but this is an important topic that raises some interesting questions. Stay tuned; I’m sure there will be much more to follow on this!

 

-Eric

Previous Post: Why I Fell in Love with the GaAs Market Again!
Leave a comment