Summary

  • Ucom's September 10 modernisation update cites 5G access for 94.6% of Armenia's population, a figure it had already reported in August.
  • Its emphasis on existing infrastructure and capacity is commercially distinct from adding coverage; earlier roaming figures illustrate changing usage, not national congestion or guaranteed returns.

A coverage map cannot show how a connection behaves when more people use it, or when the same visitors consume more data. That is the useful distinction in Ucom's latest network update. The Armenian operator is describing continuing work inside an extensive footprint, not simply announcing a new percentage of the population reached.

In its September 10 statement, Ucom says its 5G network is available in 48 cities and more than 600 settlements, as well as along major highways and at border crossings. It puts population availability at 94.6%. The company also describes phased upgrades to existing infrastructure, additional capacity and continuing performance optimisation, particularly in high-traffic areas.

The percentage is not a newly disclosed threshold. Ucom's August roaming announcement already used 94.6%. Repetition does not prove that no physical expansion occurred: the reporting method, rounding and incremental additions are not supplied. It does mean that the September figure alone cannot establish a fresh expansion milestone, let alone the quality of every customer's connection.

A bounded indication of changing load

The August release contains a more useful illustration of the difference between reach and use. Ucom says the number of foreign visitors using its network rose 54% in the second quarter of 2026 against the first, while their mobile data volume rose 93%. Traffic in that reported group grew faster than the number of visiting users.

That comparison needs its boundaries. It concerns inbound roaming, not all domestic customers or total Armenian network traffic. It is quarter-on-quarter, not year-on-year, and may reflect travel patterns or seasonality. The announcement does not establish congestion, higher revenue or a nationwide increase in usage per subscriber. Nor should inbound traffic be mixed with Ucom's separate figures for customers roaming abroad.

Still, it explains why an operator may keep investing without changing a headline reach percentage. Local demand, the timing of use and the path carrying traffic beyond a radio site can matter within an already covered area. That is an operational inference, not a claim that any particular Ucom cell is overloaded.

Three different stages of network work

Ucom's continuing Nokia relationship adds a longer timeline, not proof of a new commercial generation. Its October 14, 2025 account describes a memorandum on autonomous networks and AI applications for 6G, including joint concepts, laboratory or field trials and contributions to standardisation studies. Research cooperation of that kind is not a deployed 6G service or a demonstrated production automation result.

The same historical account says a 2024 modernisation agreement included a higher-capacity IP transport network alongside core fixed and mobile infrastructure. This is company-reported background. It helps separate physical network modernisation, current operating optimisation and future research; it does not identify September equipment purchases or their financial return.

All three cited reports come from Ucom. The reviewed material supplies no incremental September site count, investment budget, busy-hour measurements, utilisation, unit cost or revenue contribution. Its broad claims of quality and network leadership are not independent performance comparisons.

The commercial question is therefore where another unit of network spending improves service for actual usage. A new location, an upgrade at an existing site, a transport change and better operations can address different constraints. The September update points towards that work, but measuring its value requires evidence beyond the population covered.