VodafoneThree's $1.3bn core is a take-rate bet

VodafoneThree has switched on the cloud-native 5G core Ericsson built for it, the opening phase of a project the two companies price at SEK 12.5 billion ($1.3 billion). It is the carrier's latest network upgrade since its merger completed last year, and it sits inside an £11 billion ($14.68bn) investment program that has otherwise been announced in pieces, without a number attached.

Behind the upgrade sits a service only a partitioned core can deliver: earlier this month VodafoneThree launched SuperMobile 5G+, a national network slice on its 5G Standalone network that guarantees minimum download speeds of 15Mbps, with customers free to leave without a fee if it falls short. Ericsson's dual-mode platform runs 4G and 5G core functions on one cloud-native stack deployed on its Cloud Native Infrastructure Solution, which the vendor says gives the operator the foundation for the various slicing services it wants to turn on.

Capacity is the pitch. Ericsson says the core is designed for 9Tbps and claims it will be one of Europe's largest mobile packet cores, sized for streaming, gaming, and Fixed Wireless Access home broadband. Hrvoje Jerkovic, VodafoneThree's network engineering director, called the activation a major milestone in building what the company calls the UK's best network, and placed it after the integration of the radio networks earlier this year.

Underneath the launch, the vendor map is lopsided: Ericsson was contracted last year to handle rising data demand, and says it is the sole vendor for the nationwide core network while holding the significant majority of the radio access network. Nokia is still in the build: last week the two companies announced the activation of the first 100 advanced 5G radio sites on Vodafone's UK network.

A core scope with a disclosed price and one accountable vendor is the rollout's second bankable structure. When VodafoneThree's 100 new 5G sites arrived without a price, this publication argued that the eight-year Nokia RAN deal was the only bankable structure in the rollout; the $1.3 billion core contract now makes two.

The core moves the wager from build cost to take rate: nine terabits per second is capacity bought ahead of the subscribers who must fill it, and the slice being sold against it comes with an escape hatch. Capacity underwritten by a guarantee a customer can walk away from is capacity the operator carries either way. That is no argument against building it—a single core carrying both 4G and 5G traffic is cheaper to run than two, and enterprise slicing is a real market—but the return on the $1.3 billion is settled at the subscription line, not the antenna count.

That line is where it gets hard. Our reporting on Britain's 5G build noted that the sub-£20 ARPU behind the rollout is the number no planning reform touches, and a 9Tbps core with a 15Mbps speed floor is the kind of asset that has to earn its keep out of a base paying that. The number to watch is how many subscribers buy that guarantee.