This was originally posted on December 13th in the "Beyond the Next Billion" newsletter on LinkedIn and adapted here.
December 4th, 2023 saw a massive announcement of Ericsson inking a new, $14 billion deal with AT&T to become the sole supplier of radio access network (RAN) equipment. At first glance, the announcement elicited predictable headlines, mostly at the expense of Nokia, which will see its share of AT&T RAN decline down to zero. Ericsson soared, Nokia tanked and much debate ensued, including ominous equity analyst opinions about the possibility of Nokia needing to be broken up as a result of this setback.
The announcement also had explicit references to Open RAN, suggesting that AT&T had embraced Open RAN architecture and that Ericsson’s product roadmap would help AT&T “open” up its network with a view to including new vendors’ radios and other network elements. Right on cue, Fujitsu was included in the announcement, presumably for its radios to be included in an undisclosed number of sites. Other partners included in the announcement include Intel Corporation, Dell Technologies and Corning Incorporated.
AT&T, in its announcement, states that 70% of its network traffic will flow through these new, open networks by 2026, with references made to open interfaces and open APIs. Ericsson, for its part, will enable this new open, programmable network for AT&T through deployment of a number of elements - new massive MIMO radios, Ericsson RAN Compute platform, CloudRAN and the Ericsson Intelligent Automation Platform (EIAP).
All in all, this is a big win for Ericsson and certainly packs a punch in terms of the comprehensive scope of the deal. However, there are a number of issues that merit closer scrutiny.
What is AT&T’s primary motivation? There are a number of analysts based in North America that have crunched these numbers in greater detail but at $14 billion over five years, this deal doesn’t translate into a lot of CAPEX for AT&T on an annual basis, considering it previously guided at $25 billion in annual CAPEX over the same five year period. So if this is a fraction of the spend, it would be reasonable to conclude that this spend will focus on replacing the Nokia share of the RAN first, and that too, will be done incrementally. What is not clear is how AT&T will handle the costs of the swap out of Nokia equipment. Is Ericsson on the hook for financing this, or as Jay Goldberg surmises here, is there a link to onshoring production at Ericsson’s US plant which would improve “unit economics”? Indeed, AT&T’s CEO John Stankey was quoted as saying at an investor conference that the Ericsson deal would improve their “unit economics”, which would suggest that AT&T was looking for a specific level of savings from their planned network migration, and quite possibly a degree of vendor financing as well. Clearly, Ericsson was willing and able to meet their requirements, which Nokia could or would not. Stankey’s comments also offer clear clues for how they have been able to use a sharp slowdown in CAPEX spending by telcos to extract further concessions from the leading vendors.
What is Ericsson’s primary role in AT&T’s vision? It would seem that they are slotting into the position of the lead “systems integrator”, as suggested by John Baker at Mavenir. Let’s be clear, this is a great position to be in for Ericsson. They have leading products, and the AT&T network migration will give them valuable experience in deployment of Open RAN compliant networks and validate their recent embrace of the technology, a position that caught many by surprise. This support, coming from a company that was largely observing from the sidelines, does come with some significant caveats, the biggest of which seem to revolve around push for a new fronthaul interface. This could be interpreted as a self-serving move, or simply one proposed to nudge the specifications closer to their vision. Either way, there are still open questions about Ericsson’s Open RAN efforts, their compliance with specs and backward compatibility etc. Specific to AT&T, the narrative that the operator has chosen to move to a single vendor in order to migrate to Open RAN seems counter intuitive, while choosing a single vendor to also open up their network seems like a conflict of interest. Nonetheless, and assuming that Ericsson has made a number of promises to AT&T that it will keep, the deal should go a long way towards nudging the industry towards Ericsson’s product vision and to that extent, this is a big win for Ericsson.
What role, or more specifically, what share of the RAN would be relevant for Fujitsu? It is certainly heartening to see an Asian vendor grab market share in North America, something that my fellow analyst Foong King-Yew predicted here in 2021. To be fair, Fujitsu is not the first Asian vendor to acquire a foothold in North America, as Samsung Networks already achieved this with Dish Wireless and Verizon. But for Fujitsu, despite their integration with Ericsson, there are no specifics offered on scale of deployment but we believe that most of their share gains will come towards the back end of this deal as the Open RAN architecture begins to manifest across AT&T’s network.
Are things as bad for Nokia as the recent dogpile would suggest? We reckon not. There is no question that they lost this round to Ericsson, but public statements subsequent to the initial announcement from AT&T suggest that Nokia is very much a part of their CAPEX calculus going forward. This could come from Nokia radios and network elements being plugged into a future, open network based on Open RAN principles. It could also come from other parts of network CAPEX, including transport and other elements. AT&T is also quoted as saying that there are no issues with Nokia products, which could refute initial reports from a US analyst about possible issues with new Nokia products that use fan cooling. This last one certainly doesn’t seems counterintuitive since the same Nokia products are also being deployed in India, which is one of the hotter places on the planet, last we checked. Nokia’s short term prospects have certainly taken a hit, with lowered guidance for their Networks business, but there are positives. Indeed, Nokia had announcements of Open RAN deployments with NTT DoCoMo in Japan and Deutsche Telekom in Germany, sandwiched around the AT&T setback.
When seen from halfway around the world, happenings in the North American market may seem a curiosity, but there are certainly a number of telcos in this part of the world, with the possible exception of China, that will be watching the AT&T deal closely.
Asian telcos are definitely interested in Open RAN but are faced with several complications. Notably, any new Open RAN deployments must prove backwards compatibility with legacy 2G/3G networks, several of which are still in play in the region, especially in emerging Asia. Moreover, in the ultra dense urban/suburban areas that typify most Asian markets, Open RAN performance is still not proven.
That said, AT&T’s embrace of Open RAN and Ericsson’s new positioning will give a boost to the Open RAN ecosystem in Asia and make it more competitive. The opportunity to migrate to open interfaces and mix and match radios and basebands will be very attractive to Asian telcos. In the AT&T example, a Fujitsu radio with an Ericsson baseband. But it could just as well be a Nokia radio, for example, with a baseband from an emerging vendor. To enable this, integration and testing must continue apace between the major vendors like Ericsson and Nokia with the likes of Mavenir, Fujitsu and others. A new level of interoperability will give a boost to Asian telcos like Jio and Rakuten who have espoused Open RAN principles and developed their own reference architectures and roadmaps.
On the back of a sluggish year for network equipment sales in North America coupled with a slowdown in CAPEX linked to the completion of the first phase of 5G deployments, network equipment vendors are looking at multiple ways to kickstart the next wave of investments. Beyond product innovations and cost efficiencies, it looks like we are watching the beginnings of a fresh spell of vendor financing, by some if not all the vendors. This is something that large Asian telcos will watch keenly and look to leverage.