Tom Bartlett is CFO of American Tower. The company's 170,000 towers mean it is intimately involved with every major carrier in the U.S., India, and Brazil. His comments reflect the consensus views of the companies:

• 5G "relatively flat for a few years"
• Yearly "30% to 40% demand on the network" 
• Massive MIMO and multiple antennas will be used on many/most sites to increase capacity.

Bartlett doesn't see much profit for American Tower in small cells and is directing investment to other products. Despite all the lobbying and self-congratulatory politicians, I haven't seen much of an increase in small cell building or more than modest interest outside the U.S.

AMT's primary competitor, Crown Castle, is building tens of thousands of small cells but hasn't seen much effect of the much touted "we are speeding up 5G by reducing regulation" meme in DC. Outside the U.S., almost no one is doing large builds on millimetre wave with very short reach.

Perhaps the most interesting new information is that AMT has built an "Edge Data Centre" that's 20 feet square. Bartlett thinks there will be demand to put similar units at the towers from the web giants and others. The Chinese are considering going to the towers, but almost all telcos in the West have decided to place their server further back in the network.

"In Jacksonville is that we actually have built a module. It's kind of a 20 foot by 20 foot module and it's providing an on ramp to the cloud for wireline customers, not for wireless, but it's providing a capability into a small midsize type of a customer. 

And it was probably filled up within two weeks. And they're just racks in there that we're providing and we can - we picked up a co-location center in Atlanta not too long ago, which does have an on ramp to the cloud. And so we can now see where we might be able to position ourselves in this new edge compute type of facility. From Seeking Alpha transcript.

 Here are more comments from the financial call.

And that's what the unique piece about this is while it will not be kind of an overbuild, like a 4G. We think that the amount of demand on the 5G again because of speed, because a latency will require then the carriers themselves to have to put a lot of infrastructure on the site radios, antennas, MIMO [ph] those types of assets on the towers themselves to ensure that the quality of service is being received.

So it is going to look different. And it's, it's in some cases people have said well, we might see kind of with 5G, 4G again kind of being the technology of choice because there's still going to be that 30% to 40% demand on the network, right. And so the carriers themselves are going to have to add enough infrastructure on the towers to be able to support that.

But in terms of real 5G services, you might see it relatively flat for a few period - for a few years and then all of a sudden spike because of some of the use cases being able to take off, whether it's on driverless cars or whatever it might be on. But I think at least the way we're thinking about it is that like 3G, 4G is going to be with us well into the ‘20s. It's a 15 year plus type of technology.

And so the carrier whether it's, a 5G switch they can get turned on or whether it's simply just 4G a handset devices that are in people's hands, there's going to need to be more infrastructure put it - put on those sites just because of the physical limitations in terms of the existing infrastructure supporting the existing rate of demand.

So there will be ongoing requirement to amend existing leases, depending upon the band of spectrum, perhaps more densification is going to be required because the signals aren't going to be propagating to the same length, to the extent that more mid band spectrum starts to come into the mix.

So we think that there's enough complexity there to provide - I guess go back to Phil's first question. We feel good about where the market is heading and the kind of the demand that we would expect to see.