Fujitsu’s 1Finity Bets Big on India: 200 Engineers, AI-RAN and the Open RAN Waiting Game

At India Mobile Congress 2026 in New Delhi, the halls buzzed with the usual mix of 5G-Advanced promises, sovereign AI pitches, and vendor roadmaps stretching into the 2030s. But one interview cut through the noise with a rare commodity in telecom: candour. Masaaki Moribayashi, the global CEO of Fujitsu’s networking spin-off 1Finity, sat down with Moneycontrol and delivered a verdict that many in the industry have been thinking but few say out loud.

“The Open RAN itself is not booming,” he said.

That single sentence, buried in a wide-ranging conversation about India strategy, AI-RAN, and optical networking, is the kind of admission that usually gets sanitised by communications teams. Moribayashi didn’t sanitise it. He acknowledged that the market “had not expanded as quickly as expected,” even as his company doubles down on the country that many vendors now see as the next great proving ground for disaggregated networks.

The contrast is striking. Here is a company carving out more than 200 engineers for a dedicated India networking team, positioning itself for 5G-Advanced, 6G, and AI-enabled radio access networks, and yet its CEO is openly tempering expectations about the near-term Open RAN opportunity. That tension — between long-term architectural conviction and short-term commercial reality — is the defining story of Open RAN in 2026. And 1Finity’s India playbook offers one of the clearest windows into how vendors are navigating it.

Why India, and Why Now

The logic behind 1Finity’s India expansion is not hard to trace. India’s data-centre and hyperscaler ecosystem is growing at a pace that has surprised even seasoned observers, driven by AI training and inference workloads that demand massive optical capacity. Moribayashi pointed to this directly, noting that high-capacity optical networks and photonics will be critical for handling surging AI traffic. For a company with decades of DWDM line systems and transponder expertise in Japan and the US, that is a natural adjacency.

But the telecom side of the equation is more complicated. Moribayashi was blunt about the challenge: most Indian telecom operators have already selected their equipment partners for existing 5G rollouts, limiting the scope for new vendors to enter ongoing deployments. That is a polite way of saying that the incumbents — Ericsson, Nokia, Samsung — have the lion’s share of India’s 5G radio footprint locked up, and displacing them mid-rollout is not a realistic near-term proposition.

So 1Finity is looking past the current generation. The company is positioning itself for the next phase: 5G-Advanced, 6G, and AI-RAN. It is a classic long-cycle bet, the kind that telecom vendors have historically made and sometimes regretted. But there is a difference this time. The 200-plus engineers being carved out for the India networking team are not just a sales and support operation. They represent a genuine R&D and delivery commitment, one that signals 1Finity expects India to be a source of innovation, not just a market to sell into.

That matters because India has become an unlikely bellwether for Open RAN economics. The country’s operators face a uniquely brutal combination of low average revenue per user, enormous population density, and relentless data growth. If Open RAN cannot make the economic case in India, it is hard to see where it makes it at scale. Reliance Jio’s recent push on network slicing and AI-powered RAN, ahead of TRAI’s anticipated rules, is a case in point. Jio’s leadership has spoken publicly about wanting “intelligent RAN” that can learn and adapt autonomously, explicitly framing it as a way to escape the “dumb pipe” trap that has haunted operators for decades.

That is precisely the kind of operator demand that 1Finity is hoping to serve. But the timing remains uncertain.

The AI-RAN Paradox

1Finity has developed AI-RAN technology using Nvidia GPUs, joining a growing list of vendors exploring how to inject artificial intelligence directly into the radio access network. The company’s work is part of a broader industry wave that includes Nokia’s AI-RAN platform, Ericsson’s AI-powered RAN Processor, and a host of trials across Asia, Europe, and North America.

But Moribayashi’s assessment of the AI-RAN opportunity is notably measured. He said the company had yet to see clear demand for GPU-based AI-RAN deployments and was still evaluating which applications to develop and take to market. That is a significant admission from a vendor that has invested in the technology. It suggests that while the industry has largely converged on the idea that AI will play a role in future RAN architectures, the commercial use cases remain fuzzy.

The technology could eventually enable operators to use AI to optimise network performance and support new workloads, but Moribayashi was explicit that commercial adoption will depend on identifying viable applications and a clear business case. This is the same conclusion that operators and analysts have been circling for the past two years. AI-RAN is not a product; it is a capability looking for a problem worth paying for.

That does not mean the investment is misguided. The vendors that figure out the business case first will have a meaningful advantage. But it does mean that the hype cycle around AI-RAN — the breathless predictions of GPU-powered cell towers and agentic networks — is running well ahead of deployment reality. Moribayashi’s comments are a useful corrective.

The Open RAN Waiting Game

Perhaps the most revealing part of the interview was Moribayashi’s description of 1Finity’s Open RAN strategy. The company remains active in Open RAN, working with vendors such as Ericsson, Nokia, and Samsung in certain markets. That positioning — as both competitor and partner depending on the technology and deployment — is increasingly common among RAN vendors navigating a multi-vendor world.

But the CEO’s acknowledgement that the Open RAN market has not expanded as quickly as expected is significant. It aligns with a broader pattern of recalibration across the industry. The early enthusiasm of 2019-2021, when Open RAN was pitched as a revolutionary force that would break the vendor lock-in of the RAN market, has given way to a more sober assessment. Open RAN is real, it is being deployed, and it is influencing architecture decisions across the industry. But it has not displaced the incumbents at the pace its most vocal advocates predicted.

Moribayashi said he hopes the Open RAN market is “growing from now.” That phrasing — hope rather than expectation — is telling. It reflects a vendor that has made a strategic commitment to disaggregation but is not betting the company on a rapid inflection.

There are signs that the market is indeed maturing. The OCUDU Ecosystem Foundation’s recent release, OCUDU 26.10, brought satellite connectivity and new 5G capabilities to open source RAN software for CU/DU, including 3GPP Release 17 NTN support and O-RAN Split 7.2b flexibility. Meanwhile, WiSig Networks, a Hyderabad-based company incubated at IIT Hyderabad, has opened commercial orders for an indigenous 32TR Massive MIMO radio designed for 5G networks using the n78 mid-band frequency, supporting Class A and Class B uplink performance improvement under the O-RAN Alliance’s 7.2x standard. These are incremental but meaningful steps toward a more diverse and competitive RAN supply chain.

Yet the commercial momentum remains uneven. Nokia’s AI-RAN push in India, with commercial deployments potentially beginning around 2028, is a reminder that the next phase of RAN evolution is still several years away. And the incumbents continue to dominate the installed base.

The Optical Opportunity

While the RAN story is one of patient waiting, the optical networking opportunity in India is more immediate. Moribayashi’s emphasis on optical networks and photonics reflects a reality that is becoming impossible to ignore: AI is a bandwidth problem before it is anything else.

India’s data-centre buildout is accelerating, driven by hyperscalers and AI service providers who need massive amounts of capacity between compute clusters. The optical layer — DWDM line systems, transponders, and the photonic infrastructure that carries traffic across long distances — is where much of that demand will be met. 1Finity’s long history in optical networking, particularly in Japan and the US, gives it a credible position to compete for that business.

The company’s decision to build out a 200-plus engineer team in India is partly about serving that optical demand. It is also about establishing a presence in a market that is expected to grow for years to come. Moribayashi said the broader telecom market would continue growing as AI adoption increases the need for network capacity, reliability, and low latency. He added that mobile networks would remain important for emerging applications such as physical AI and edge connectivity.

That is a balanced view. Optical is the near-term opportunity; mobile RAN is the long-term bet.

What It Means for Open RAN

1Finity’s India strategy is a useful case study in how vendors are adapting to the current phase of the Open RAN cycle. The company is not retreating from Open RAN, but it is not betting everything on a rapid market expansion either. Instead, it is diversifying across optical networking, AI-RAN research, and next-generation mobile standards, while maintaining a presence in Open RAN through partnerships with the major RAN vendors.

That approach may not generate headlines, but it is pragmatic. The Open RAN market is growing, just not as fast as the most optimistic forecasts suggested. Operators are deploying disaggregated networks in select scenarios, but the economics remain challenging in many markets. The vendors that will succeed are those that can sustain investment through the slower-than-hoped ramp while positioning for the next wave of demand.

Moribayashi’s comments also highlight a broader truth about AI-RAN: the technology is not yet a commercial product. It is a research direction with promising potential but an unclear path to revenue. The vendors that figure out the business case will be the ones that win. Until then, the industry will continue to invest, experiment, and wait.

For India, the stakes are high. The country has become a test case for whether Open RAN can deliver on its promise of lower costs and greater flexibility in a market defined by extreme scale and price sensitivity. If 1Finity’s bet pays off, it will be because the company positioned itself for the next generation of networks before the market was ready. If it doesn’t, it will be another data point in the long, uneven story of Open RAN’s commercialisation.

Either way, the waiting game continues.

Sources

• Fujitsu’s 1Finity eyes India expansion as AI, data centre boom drives optical networking demand: Global CEO
• Fujitsu’s 1Finity Goes Big on India: 200+ Engineers, AI-RAN & Data Centre Boom | IMC 2026
• AI-RAN India: Nokia Plans 5G Shift by 2028
• Jio Deploys 5G Network Slicing Ahead of TRAI Rules
• OCUDU 26.10 brings satellite connectivity, 5G to open source RAN
• Indigenous 5G Radio Opens Orders in India