Open RAN 'by default?' Not bloody likelyOpen RAN 'by default?' Not bloody likely

The mutant varietal of "single vendor open RAN" now infests telecom, strangling at birth the crossbreeds the industry would spawn by coupling vendors at the same mobile site. Cheerleaders claim not to be scared, insisting "single vendor" is no dominant species wiping out others but the genesis of a more competitive and vibrant market. As radio access network (RAN) products become increasingly based on open interfaces, truly multivendor networks can take root, they say. Encouraged, a few commentators have begun using "open RAN by default" instead of "single vendor open RAN" as the signpost to this fertile ground. Enthusiasts should be wary.

As they typically do, open RAN's biggest advocates have spent so long peering down the microscope at the technological nitty-gritty that they have lost sight of the economic and market realities. Open RAN obviously needs those non-proprietary interfaces to succeed at scale, and these are more developed than they were just a couple of years ago. It is technologically easier to combine supplier A's baseband with supplier B's radios than it ever was. But that does not mean it will be what operators mainly do.

The number one problem for genuinely multivendor open RAN is a shrinking market of barely profitable vendors. RAN product sales worldwide generated revenues of about $45 billion in 2022, says Omdia, an analyst company owned by Informa, Light Reading's parent. In 2023, sales fell to $40 billion and, after a disappointing first nine months of 2024, they are expected to land at roughly $35 billion for last year, according to the midpoint of Omdia's most recent forecast. Jobs are being scrapped like decorations after Christmas as the likes of Ericsson, Nokia and Samsung – the biggest kit makers outside China – take steps to protect their margins.

These are already wafer thin at Nokia after a loss of US market share. Yet to publish full-year results, it reported a third-quarter operating profit of just €92 million (US$94.7 million) on sales of €1.75 billion ($1.8 billion) at its mobile networks business group. Mavenir, the biggest of the US open RAN challengers, has also been cutting jobs after ratings agencies warned it was at risk of default this month. John Baker, Mavenir's top spokesperson for open RAN, recently left the company, as revealed by Light Reading in December. These are not and never have been the economic and market conditions in which new entrants typically thrive. They are, in fact, more likely to result in consolidation.

Not a Kodak moment

The telco desire for a more competitive supplier market and lower-cost products is understandable. Their own sales have flatlined after a smartphone data boom and 5G, the latest mobile technology, has largely failed to spur growth. But openness does not change the underlying technology, which is based on the same 4G and 5G standards found with proprietary interfaces. Open RAN is not disruptive in the way cars were to horse-drawn carriages, or digital photography was to film.

The only cost savings available to a telco are from lower prices offered by new competitors determined to build market share. Yet price wars are usually won by the biggest companies with the deepest pockets (Reliance Jio, a telco new entrant, managed it in India only because it is owned by one of the country's largest conglomerates, and it also received a helpful shove from regulators hostile to foreign investors).

Japan's NTT Docomo, perhaps responsible for the world's most extensive open RAN deployment so far (although critics say it is not based on industry-wide specifications), has now acknowledged the technology does not bring any major economic benefits. "It is competitive in most cases," said a spokesperson for OREX SAI, a Docomo subsidiary investing in open RAN products, when asked how the total cost of ownership (TCO) compares with traditional RAN. Answering questions by email, the company spokesperson added that "our TCO is competitive to that of incumbent vendors."

New entrants could always dislodge incumbents by concentrating their research and development (R&D) on specific areas. This was, after all, the original premise. In a world of closed interfaces, a vendor needed to bring a pre-integrated set of products to any 5G tender. With openness, it could run its software over another company's hardware or link its radios to someone else's baseband. But this is not how the market has evolved.

The continued preference of Tier 1 telcos for dealing with a relatively small number of suppliers, if not a single-vendor open RAN, has compelled prominent specialists to change course. Mavenir, the best example of this, now straddles hardware, software and systems integration, having started out purely in software. Accordingly, it is effectively up against Ericsson and Nokia on their own terms as a vendor of numerous RAN products. But the Nordic vendors are thought to spend about $5 billion on RAN R&D each year, a figure equal to about 19% of their collective RAN sales. When it last shared numbers, Mavenir had generated total sales of about $500 million for 2022, including $100 million in the RAN market. Without a super-generous and uber-wealthy investor that it currently lacks, Mavenir cannot possibly hope to compete.

Any success for Mavenir or another challenger in a shrinking market would also have to come at the expense of incumbents that already look weak. Ericsson and Nokia have put guardrails around R&D budgets, despite making cuts elsewhere. But those guardrails would inevitably start to wobble if market share were lost. A reduction in spending by 5G's main developers outside China would conceivably hinder innovation and force up unit prices.

The elephant in the room

Open RAN "by default" is true in the sense that specifications are increasingly baked into products by vendors big and small. Nokia made no reference to open RAN in its statement last year about a major 5G contract with Vodafone Idea. But the products it is supplying to India's third-biggest telco are compatible with specifications, according to a reliable industry source.

What's dubious is whether this compatibility really matters if economic and market realities effectively limit the opportunity for multivendor deployments. Docomo's rationale for creating OREX SAI is partly that few operators have the expertise or resources to invest in the systems integration needed to stitch together different vendors. Unless someone like OREX SAI does the job, many telcos will continue to rely on Ericsson and Nokia for this activity. But in the absence of commercially viable smaller vendors, the only option for telcos will be combos of giants. It is hard to imagine fierce rivals like Ericsson and Nokia being prepared to collaborate so much. It is equally hard to see how such a partnership would suit telcos calling for more supplier diversity.

Meanwhile, the elephant in the room is Ericsson, with its half-heartedness about open RAN. After landing a $14 billion deal with AT&T about a year ago, it has seemingly convinced former skeptics of its full commitment to the approach. Yet it admitted as recently as mid-November that its 6672 processor, described in late 2023 as a groundbreaking 5G technology, does not and will not support open RAN for massive MIMO, an advanced 5G system. However open RAN is defined, "by default" sounds very wide of the mark.