It's an interesting time in telecommunications because the fundamental approaches to carrying data are changing in dramatic ways.

Two changes are particularly impactful: virtualization and 5G. Virtualization is the core concept behind software-defined networking (SDN) and network functions virtualization (NFV), both of which are massively impactful. The other great change is in wireless. The industry is on the cusp of the 5G plunge, which will have massive ramifications as well.

In times of such fast-paced change, it is important to look at how various players are spending their money. Some key questions: How much is being spent? Where is it being spent? Where is it not being spent? What type of company is spending more? What type of company is spending less? Are companies spending in a way that suggests they believe a dramatic change in networking is near?

The impact of the cloud, which goes hand in hand with SDN and NFV, is summed up quite well by Eric Jhonsa at the end of a roundup on spending patterns at Real Money:

"Look for fresh signs of this still-widening chasm between the spending habits of telco/pay-TV service providers and cloud giants as earnings season progresses. The first group continues facing major top-line pressures, and is spending very cautiously because of it. The second group is seeing much better business conditions, and is comfortable making giant investments both to roll out new services and meet the considerable infrastructure needs of existing ones."

Jhonsa is not the only one who sees the trend. IDC sees a continuing movement of money to the cloud. Last month, the firm said that spending last year on cloud servers, enterprise storage and Ethernet switches would be determined to be $46.5 billion. That would represent a year-over-year increase of 20.9% compared to 2016. This will be divided between public cloud data center and on- and off-premise private cloud environments. IDC said spending on traditional infrastructure will decline by 2.6%. Though legacy spending still will be greater, the gap is shrinking, as it has for at least the past seven years, according to a graphic with the press release.

5G, which will start being deployed in earnest during the next couple of years, portends great changes as well. FierceWireless reported, for instance, that Sprint is issuing a request for proposals for small cells and towers. 5G is a milestone because it creates a rough equivalency between wireless and wired networks. This gives people who build, repair or extend networks a potent and generally less expensive option, since sending signals through the air in most cases is cheaper than digging up streets to install fiber.

Let's look at how these changes could impact capital expenditures.

Transitioning from a legacy to a virtualized network is dramatic. Its impact will be felt on fiber deployments, the nature of the equipment riding that equipment, the tools that manage the network and the training given to operational personnel.

For instance, late last month AT&T announced the disaggregated network operating system (dNOS). One of the attractive features of virtualization is that less expensive generic equipment can be used. The differentiating elements - what makes a router a router and a firewall a firewall, for instance - are embedded in software that is downloaded to them. AT&T is working with the Linux Foundation to create an operating system capable of supporting these cheaper and far more uniform devices.

The basic theme of dramatic change is much the same in the world of 5G. In most scenarios, this approach to wireless uses very high frequency multi-millimeter (mmWave) frequencies. These require a layer of "small cell" antennas to be inserted between customer equipment and today's towers. That's a massive investment.

All of these changes impact the amount of money that is invested. The bottom line is that networking is changing radically. Carriers and enterprise planning executives must make decisions that are on a grander scale than decisions made when the pace of technology change is not as extreme. Part of this is knowing when to close the checkbook on technologies that are on the verge of obsolescence.

The changes are not only technical. The end of the net neutrality rules late last year likely will improve prospects for the big carriers by enabling them to more fully dominate data distribution. For instance, innovative programming and services - generally short-handed as the "next Facebook" - are more likely to come from an entity with ownership or other ties to the carrier.

One constant would remain: New services would generate revenue for carriers in their roles as ISPs. Now, however, these dominant companies are more likely to benefit from the content itself. That does two things: it makes it more likely that they will increase capex and it slightly changes where the money is spent due to an overall simplification of networking that the end of net neutrality would bring.

The bottom line is that technology is changing faster than ever. This, combined with legal changes, is making technical and financial forecasting and investment a challenge - but one that is not necessarily unpleasant for planners, since the result of good choices is increased revenue and profits.