Or: why a market leaver became a market leaver. Private 5G is growing fast, but not explosively. SNS explains the market maths, the hype gap, and why Nokia has called time on a critical value prop for the new physical AI economy.
In sum – what to know:
Market forecast – Private 5G is forecast to grow from $2.8bn today to $6.6bn by 2029, but the market is a long-term industrial play rather than an overnight sensation.
Telecoms engine – SNS says private 5G will be a major growth engine for telecoms, even as public 5G infrastructure stagnates and profitability varies by deployment.
Nokia tension – Finish vendor’s decision to sell its Enterprise Campus Edge unit highlights the tension between private 5G’s strategic importance and difficult economics.
Picking up again on this private 5G discussion with SNS Telecom & Research last week, which exploded a market forecast, about $6.6 billion of value by 2030, into an editorial thesis about private 5G as a blueprint for AI-era telco infrastructure, and subsequently went to town on physical AI as the demand-side catalyst to pull private 5G into the mainstream. There were more questions besides, and plenty of leftovers to cover from the original SNS briefing. We may come back to these other drivers and impacts, but we should deal with some of those questions.
These include straightforward ones, just to frame the forecast, plus reflections on the hope and hype in the sector, plus a longer conversation about the hue-and-cry with Nokia, the erstwhile leader in the space, which issued an update with its Q2 earnings last week that its Enterprise Campus Edge (ECE) division – which went a long way to build the market in the first place – has been reclassified among its newly “discontinued operations”, and should be sold by the end of Q4. Asad Khan, research director at SNS, is on hand with all the answers.
So firstly, that $6.6 billion forecast, by the end of 2029 – what is the startpoint for private cellular networks, as of now (2026)? The answer, missing from the original press note, is $2.8 billion, it turns out. Which is information RCR did not have before, entirely – which says the ‘private 5G’ market (if we can just call it that for SEO) is to more than double (plus-135 percent) in three-to-four years, depending on the start date in 2026. Actually, we can work it out: SNS said last time the growth forecast is 34 percent per annum – which is a four-year CAGR, from the start of 2026.
There’s some further (envelope) maths in the last piece, specifically comparing the SNS figures with forecasts from Dell’Oro Group, including for RAN sales in the private, public/global, and GPU/AI sectors. The point is that, whether at SNS’s top-end run-rate of 34 percent or Dell’Oro’s lower (and looser) RAN forecast, at 10-20 percent, private 5G will outrun practically every other part of the mobile industry for growth over the next few years. So how big is it, actually? “Currently, [it] represents just over seven percent of total cellular infrastructure spending,” says Khan.
He adds: “We expect it to increase to as much as a third by the mid-2030s.” So a few years, and longer then for that spiralling CAGR forecast.” And how profitable is it? He responds: “The economics of wide-area private 5G is similar to public 5G, where the vendor typically sells directly to the end user responsible for operating the network. The margins can come under pressure in the campus segment, particularly when larger vendors supply smaller deployments. The tech works and the demand is growing, but profitability varies significantly project-to-project.”
Which perhaps explains Nokia’s exit strategy – too many customers, too much work, too little (deterministic) profit. Which is funny because Ericsson does not think like that (yet anyway), and because private 5G, as discussed, might be viewed as a fertile edge function in Nokia’s so-called ‘AI supercycle’ discipline. We will get back to Nokia. First, let us consider what it is quitting (and the term might be debated, as below) with the sale of its ECE unit. How important is this market? “Very important,” says Khan. The public 5G market is flat, he says; this is where growth can be found.
Khan explains: “The public 5G infrastructure market – and the $30-45 billion RAN market, in particular – is likely to remain stagnant, aside from the usual investment cycle peaks we’ll see with 6G from 2029/30, or barring something like Starlink investing in terrestrial RAN infrastructure in the US – which is likely to benefit a single vendor. The private 5G market won’t replace or become as big as the public cellular business overnight, but it is where most vendors see steady growth over the coming decade and beyond.”
He adds: “Think of the thousands of factories, warehouses, mines, ports, airports, and hospitals that need reliable wireless coverage for physical AI and connected workers, rapidly deployable 5G systems for armed forces worldwide through initiatives such as the UK’s Project Morpheus or Germany’s D-LBO (digitalization of land-based operations), nationwide (or statewide in the US) private networks for utility comms, and the 210,000 kilometers of railway tracks where GSM-R will be replaced by 5G-based FRMCS networks.”
Note here, that “thousands” figure was indexed once, seven years ago, by Nokia (via Boston Consulting) as 14.5 million different venues – and mostly campus-sized venues or smaller. And Khan neatly splits his rundown, here, between smaller campus sites (“factories, warehouses, mines, ports, airports, and hospitals”), which are not clear candidates for post-ECE Nokia core network systems, and larger industrial venues (utilities and railways) that will take more lightly-modified versions of Nokia’s traditional RAN and core products. Military venues will take both.
There is an argument that Nokia’s 2019 market-sizing drove expectations through the roof, and made its own ECE unit a hostage to outrageous fortune. Is the market over-hyped? Or is the hype justified? “Well, it’s been a consistently-growing market since we started tracking it in 2016, but there have certainly been periods when it was overhyped – particularly around 2020 – which led to unrealistic growth expectations among some stakeholders. Our view is the market will continue to grow over the coming decade, but it will do so steadily, rather than explosively.”
Khan adds: “Every deployment has its own business case and operational requirements, so this is always going to be a long-term growth story rather than one driven by hype.”
Okay, so let’s deal with Nokia. Here’s a quick Q&A with Khan, asked directly about the Finnish vendor’s strategy in the private 5G market, and whether SNS thinks it makes good sense, or bad. RCR (or me, anyway) has written enough on the topic, and will withhold any further editorialising here. You can make up your own mind. Note, Khan’s answers were provided to RCR before Nokia’s statement last week that it expected to sell its ECE division by the end of 2026, having already agreed a parallel deal with Inseego for its FWA unit.
– Who might buy the ECE unit?
“At this stage, it’s pure speculation, [but] the most likely scenario is either another vendor or even a global systems integrator seeking to capitalize on ECE’s technical expertise and relationships with over 200 channel partners worldwide. There is still a very slim possibility that Nokia ultimately decides to retain the ECE business unit or spin it off as an independent entity. We’re expecting a decision by Q4 this year.”
– What does it mean for Nokia?
“Unless ECE is acquired by a rival RAN vendor, Nokia is likely to continue to supply its RAN equipment to campus network customers through ECE’s new owners. Within Nokia, there will be more of a focus on end-to-end solutions for wide-area private networks (and select mission-critical or large-scale campus network deployments), [and its] AI-RAN ambitions with NVIDIA and desire to win back share in the public RAN market through 6G network deals starting in the 2029/30 timeframe.”
– Is the strategy good, or not?
“It is too early to call it either a success or a mistake, but the timing is certainly surprising given the recent momentum behind larger and multi-site private 5G projects in the campus network segment. The move may make sense from a short-term profitability perspective, but it’s much harder to understand in light of the market’s direction and the more than decade-long effort that has gone into building the ECE business unit.”
– What does it mean for ECE?
“As a standalone unit, ECE is continuing to add new features to its flagship DAC (Digital Automation Cloud) private wireless solution, including a new all-in-one small cell, an updated compact private wireless system capable of supporting up to 12 radio nodes and neutral host capabilities. The RAN equipment that ECE supplies as part of its DAC solution will almost certainly remain under Nokia’s ownership, while the compact core and other DAC software assets could form part of any transaction, depending on how the sale is ultimately structured.”
– What does it mean for customers?
“For existing customers, there shouldn’t be any immediate disruption. Support contracts and product development are expected to continue under the new owner. Some of the most experienced industrial enterprises are already accustomed to this type of supplier transition. A decade ago, Alcatel-Lucent’s early private LTE customers, including Rio Tinto, experienced a similar change when Nokia acquired its French rival. The bigger impact is likely to be on prospective customers of Nokia’s equipment, some of whom may choose to delay investment decisions until there’s greater clarity around ownership, product roadmaps, and long-term support commitments.”
– Will Nokia be market-leader in 18 months?
“In terms of vendor share, Nokia currently accounts for approximately 28 percent and 22 percent of global private cellular RAN and core sales respectively – slightly behind Huawei due to the latter’s dominance in China’s well-established private 5G market. If Nokia continues supplying the underlying RAN infrastructure to ECE under its new owner, the commercial impact on Nokia’s market position is unlikely to be as severe as many people initially expected. In the next 18 months, we still expect Nokia to remain one of the top three suppliers globally and one of the top two outside China. Any loss of share will likely benefit specialist private wireless vendors and Ericsson.”
– Who will benefit most from the break-up?
“Specialist RAN and core vendors – such as Celona, Airspan, Globalstar, Siemens, Druid Software, RADTONICS, Pente, Ataya, GuardStack/Blackned, Cumucore, AttoCore, Saviah, HPE, Cisco, Star Solutions, Benetel, Sercomm, Askey, HFR, and many others – will benefit the most. We are already seeing early signs of this, with the likes of Celona winning multi-site campus projects for customers including Cargill, BP, LyondellBasell, Marathon Petroleum, Celanese, and Google (for data centers).
“That would have been much harder to imagine just a few years ago. We also expect to see an increase in multi-vendor deployments, with Nokia’s RAN being paired with mobile cores from suppliers such as HPE, Cisco, and Druid Software. One recent example is Sydney Metro West’s FRMCS-ready private 5G network, where BAI Comms is deploying Nokia’s RAN alongside an HPE core. In previous rail sector deployments, Nokia would almost certainly have supplied an end-to-end solution. The uncertainty surrounding the proposed transaction has also created an opportunity for Ericsson. W’’re already seeing Ericsson work much more closely with integrators like Future Technologies – which has been a strong Nokia partner historically.”
