We called the 5G-to-6G lull in our 4Q24 report, published in February 2025. Eighteen months on, it is still the most accurate one-sentence description of this market. The 2Q26 numbers do not change that. What they do is show where the money is moving while the industry waits.
Total RAN and core network revenue rose 3% in 2Q26, both year-over-year and sequentially. In a dormant market, the composition of that 3% is more interesting than the 3%.
The quarter
5G RAN carried part of it. 5G Core carried more, growing 40% sequentially and 20% year-over-year. And EPC/vEPC grew again — which surprises people until they look at the installed base. There were 297 5G non-standalone networks live as of June 30, and NSA still predominates, so capacity upgrades keep flowing into a core generation the industry has been describing as legacy for six years. Let’s face reality: 400+ cellular networks worldwide are still on plain LTE and technically that is not 4G!
Standalone continues to arrive slowly. Three new 5G SA networks went live during the quarter, bringing the total to 95, up 30% year-over-year. That is real growth off a small base, and it remains the gating factor on most of what 5G was supposed to enable.
Geographically, the growth came from Northeast Asia, led by Japan; Southeast Asia, led by Vietnam; Australia and India; and EMEA. China added 120,000 5G base transceiver stations in the quarter, taking its 5G footprint to 5,102,000.
Vendor ranking, and the number underneath it
RAN sales: Huawei, Ericsson, Nokia, ZTE, Samsung. Core sales: Huawei, Ericsson, ZTE, Nokia. Quasi-static, as it has been for several quarters.
The more useful cut is what happens when you remove China. Ex-China, the core ranking inverts to Ericsson first, Nokia second, Huawei third. China's 5G connection footprint is doing the work in the headline numbers for both Huawei and ZTE, and any strategic read of vendor position that ignores that is measuring the wrong market.
On 5G SA core commercial network footprint, Ericsson and Nokia remain neck and neck.
The 2026 outlook: RAN down, core up
We expect RAN slightly down year-over-year for 2026 and 5G Core up. The activity is upgrades rather than builds — 5G-Advanced, rip-and-replace at Bell, Telus and Deutsche Telekom, and expansion at Rakuten Mobile, much of it carrying open RAN, vRAN or cloud RAN components.
Global telecom capex remains murky, and the capex-to-opex shift is accelerating. CSP software spend accounted for 10% of total capex in 2024; we’re still missing a few numbers to make the call for 2025.
Which brings us to the number we would put in front of any vendor CFO reading this: Rakuten Symphony reported RAN software at 93.5% of its total RAN revenue. That is one operator with an unusual architecture, and it is not the market. It is also the direction of travel. Hardware's share of total RAN sales will keep falling as software's rises, and the vendors whose margin structure assumes otherwise have a problem that no 6G cycle will solve for them.
Where our calls stand
Open vRAN past half of all RAN by 2031 — on track. We have it at 55% in 2031, from 9% in 2Q26. That is the single largest structural shift in this forecast and the one we get the most pushback on. Nine percent today is why.
4% CAGR, 2026–2031, with a 6G surge beginning in 2030 — unchanged. Our 20-year CSP wireless infrastructure footprint pattern analysis continues to point here. 5G stays flattish through 2027 and begins declining in 2028.
No 6G core — unchanged. For reasons set out in our 6G Infrastructure Outlook, we do not expect a 6G core to be developed. We are aware this is the least consensus position we hold. We are holding it.
5G Core at a 10% five-year CAGR against EPC at -43% — on track. The drivers remain fixed wireless access, still the undeniable success story of this generation, followed by VoNR and network slicing. RedCap and non-terrestrial networks are the emerging additions.
Téral Research's View
One. The lull is not a pause before a return to the old pattern. RAN revenue declining from 2028 while open vRAN climbs to 55% means the 6G surge in 2030 arrives into a market with a different cost structure than the one 5G was sold into. Plan for the 2030 revenue, not the 2030 unit volumes.
Two. If you build core networks, the 43% EPC decline and the 10% 5G Core CAGR are the same sentence read twice. The installed base is still paying for capacity today — 297 NSA networks are not upgrading themselves — but that revenue has a visible end date and the replacement is growing at a quarter of the speed it's disappearing.
Three. Take the ex-China vendor ranking seriously. Two very different markets are being averaged into one league table, and the competitive dynamics in each bear almost no resemblance to the other.
We have now spent six quarters describing this market as dormant. Dormant is not the same as static. Underneath a 3% quarter, software is taking RAN, standalone core is growing at ten times the rate of the thing it replaces, and the largest architectural transition in the industry's history is sitting at 9% and climbing.
The full report, "2Q26 Wireless Infrastructure Market Report," including quarterly market size, vendor market shares, and five-year forecasts by region and technology, is available to Téral Research clients. For access or a briefing, contact us at contact@teralresearch.com.