Parallel Wireless, one of the founder companies of the Open RAN movement, has had to bow to market conditions and significantly scale back its operations. TMN has heard that up to 80% of the company’s payroll – so perhaps 700-800 of 1,000 people – may have been let go, although the company’s funder, founder and CEO Steve Papa has not confirmed that number, despite attempts to request comment so far.
The news has been interpreted as an Open RAN story, but perhaps it is just as much a “new company in a very entrenched market” story. Parallel hasn’t always been an “Open RAN” company, but it has always had the ambition of transforming the way mobile networks are built and deployed. When it launched it outlined a vision of a company that enabled wireless networks to be installed and operated as simply as enterprise WiFi. Papa’s vision was to provide low cost cellular equipment that was cheap to build and simple to deploy and manage. It worked its way through various scenarios, from small cells and enterprise coverage solutions to providing full macro coverage.
Allied to those ambition was its early move to an architecture that enabled it to take advantage of virtualised network functions. Its HetNet Gateway was an orchestrator node that leant into the disaggregated RAN model, with the abstraction of RAN intelligence away from the radio-antenna unit. It was also an early explorer of the potential for containerised, cloud-native RAN functions – and in fact had engaged with Robin.io on cloud management and automation of workloads well before Rakuten CTO Tareq Amin found the company via a Google search, and plugged it into the Rakuten ecosystem.
Through its engagement with TIP Open Cellular programmes, and then the O-RAN Alliance specifications, Parallel’s initial vision morphed into a more explicit Open RAN position, as Open RAN networks provided a potential route to market for a company intent on disrupting the sales cycles of the major vendors.
Its architectural model was also to fit nicely into evolving Open RAN specs, with the Het Net gateway aligning with the emergence of containerised cloud-native RAN functions, and the development of near real time RIC as an intelligent controller in the RAN.
All of this points to a company that made prescient bets on the direction of the industry. But as it made that journey, Parallel and its CEO Papa identified some potential hurdles for any company active in the Open RAN space.
First, a RAN software company that was trying to act in a disaggregated network either needed the existing Radio Unit vendors to open up and connect to a third party baseband, or it needed to see the rapid emergence of a new cadre of open RU providers. Without this, 5G deployments from new vendors would exist as a separate layer from the 4G network, or would be limited to greenfield and edge cases where existing 4G was not deployed already, such as indoors or rural. Parallel was an early advocate for the incumbents opening up their interfaces – something which they continued to drag their feet on.
Papa also saw that the established vendors – and especially Huawei – had a long head start in developing RUs with the power and performance profiles MNOs were looking for. He called for more investment in radio silicon, so that there could be a vendor ecosystem that could compete with the major vendors in this area – especially with Huawei’s Gan-based radio-antenna units for massive MIMO. Papa wrote, in 2020, “In order for operators to make big ORAN investments they need confidence that Open RAN can solve their entire problem and it is critical ORAN has an answer for 5G massive MIMO. To become competitive with Huawei’s 5G, the place the industry needs to put the money is in the semiconductor innovation that enables Massive MIMO radios that are cost effective and more energy efficient than Huawei. If Ericsson and Nokia are struggling to be competitive with Huawei’s radios we should not expect ORAN to magically solve this problem by using the same semiconductors available to Ericsson and Nokia at present.”
The company also identified early on that Open RAN could not just provide a 4/5G overlay for operators. 2G and 3G support would be required for years and an operator investing in 4G and then 5G would not want to support multiple radio layers from different vendors. Single RAN solutions were the answer, and the virtualised RAN providers would have to match this. Parallel articulated an all-G vision that had some cut-through with operators. One trial in Turkey removed an existing vendor’s entire stack and replaced it with Parallel’s multi radio. Later, Mavenir moved to acquire ip.access partly to acquire its proven 2/3G radio stack. Even later still, Samsung has just this week announced that it will support 2G from its vRAN before the end of 2022.
The company also positioned itself explicitly as a US company when that was politically expedient, hoping to attract the attention of operators and legislators trapped between a reliance on Chinese vendors and a Western duopoly.
In latter years its marketing team made recurrent claims of progress in many national markets, and its website says it has 60 operator customers – but these seemed not to materialise into major announcements. Altiostar, now part of Rakuten Symphony, made the initial big Open RAN deal with Rakuten. It and Mavenir then won the initial Open RAN share at Dish. Samsung followed in behind both of those two with a billion dollar contract at Dish. And Samsung was also the winner of the high profile Vodafone UK deal to strip out Huawei kit in 2,500 rural and semi-rural sites.
Parallel won deals with MTN and Vodacom in Africa, in the Middle East with Zain and Etisalat, and with Vodafone in Turkey and in South America with Telefonica, along with some local/regional players in North America. But the really “big one” never dropped, despite all this promise.
In response to these challenges, Parallel’s marketing evolved to long series of articles and posts explaining Open RAN architecture and technology, spelling out these issues and the steps that were required to overcome them. These efforts saw it act as a “market maker” but not, it seems, a rain maker. Nevertheless, the company said as recently as Q3 2021 that it wanted to double its staff numbers before the end of 2022. Anecdotally, a staffer recently told TMN that one of the main problems the company was facing was recruiting and scaling up, and LinkedIn remains full of job adverts for the company. The decision to turn the tap off seems to have been a sudden one.
The reasons for that are, to an extent, out of its control. Yes there was covid, the chip shortage, but also the fact of pure timing and the ways that large MNOs procure network technology. In mature markets, operators were under pressure to get 5G deployed quickly, often for strategic national prestige reasons, or just for competitive advantage. Accordingly, they were always likely to look to mature vendors and technology to provide the solutions they needed, rather than go all-in on new vendors. That’s not even about Open RAN, purely a question of, “Which vendor will get me to 5G in the most performant, risk-free and quickest way possible?”
Perhaps if Phase One 5G had been a 2025 play, things might have looked different, but there simply wasn’t enough time or motivation for large scale operators to move their macro 5G roll outs to new vendors. Even those mature operators that have engaged heavily on Open RAN messaging have written their largest cheques to Samsung.
Outside of the mature markets and 5G, Parallel was competing against ZTE and Huawei, which could provide very low cost and export credit-guaranteed terms to operators. So there was a squeeze – and it’s one faced by all Open RAN companies at this point. In emerging markets, you need to meet cost profiles that are very difficult for early stage companies. In mature markets, integration with incumbents is a huge hurdle, and those same incumbents have a huge advantage in being able to upgrade their 4G deployments to 5G.
In technology development, architectural choice, in identifying the key market needs, Parallel Wireless got a lot right. Yet it couldn’t convert its ambitions, and its vision, into sustainable business success. The questions for the rest of the Open RAN vendors are these: are they similarly challenged, and if so do they have the capacity, the money and the presence to face down the issues that have reduced Parallel its current position?