The predictable and immensely profitable decennial cycle that has defined the mobile industry for generations, a rhythm that vendors once likened to a toymaker awaiting Christmas, appears to be irrevocably broken. For decades, the arrival of a new “G” signaled a massive, global spending spree on new network hardware, creating a reliable revenue boom for equipment suppliers. However, the commercial disappointment of 5G has shattered this long-standing model, leaving behind a trail of disillusioned telecom operators who invested billions in capital for a network that failed to deliver any meaningful growth in sales. This profound financial hangover has not only eroded the industry’s appetite for another hardware-centric spending bonanza for the forthcoming 6G standard but has also triggered a fundamental, and painful, reinvention of how mobile networks evolve and the companies that build them. The era of distinct, revolutionary hardware generations is giving way to a future defined by continuous, software-driven evolution.
The Operator Rebellion Against Hardware
The core of the issue stems from 5G’s inability to deliver a “killer app” or a compelling new service that could persuade consumers and enterprise customers to spend more on their connectivity. Unlike 4G, which unlocked the mobile app economy, or 3G, which brought the internet to phones, 5G’s primary benefit was simply providing much-needed additional capacity for networks already congested with data traffic. While essential, this function did not create new revenue streams for operators. They were left with the massive bill for spectrum auctions and nationwide equipment rollouts with no corresponding return on investment, a situation that has made them deeply skeptical of repeating the process. This financial strain has created a powerful consensus among the world’s largest mobile operators: the old way of doing business is no longer acceptable, and the next wave of network upgrades must not follow the same expensive, hardware-dependent path that characterized the 5G deployment.
This sentiment has crystallized into a formidable, unified front against another hardware-driven refresh cycle. The Next Generation Mobile Networks Alliance (NGMN), an influential consortium representing major global operators, has unequivocally stated in a white paper that “6G must not inherently trigger a hardware refresh of 5G RAN infrastructure.” Instead, the alliance is pushing for 6G’s capabilities to be introduced primarily through software-based feature upgrades on existing network elements. This position is not merely corporate policy; it is a sentiment echoed by top executives across the industry, effectively serving as a direct mandate to equipment vendors. The message is clear: the future of network evolution must be more cost-effective, sustainable, and less disruptive. This operator-led demand for a software-based evolution over a hardware-based revolution is the primary catalyst dismantling the traditional “G” cycle and forcing vendors into a radical strategic pivot.
A Fundamental Restructuring of the Vendor Landscape
In response to this immense pressure from their biggest customers, major Radio Access Network (RAN) vendors are being forced to abandon what was once their most reliable and lucrative business: the design and sale of proprietary hardware systems built around expensive, custom-developed silicon. The development of these application-specific integrated circuits (ASICs) for baseband processing is an incredibly costly endeavor, with research and development spending soaring in recent years. For example, Ericsson’s R&D expenditure climbed from 14% of revenues in 2016 to nearly 22% recently. In an era where operators refuse to pay for a full-scale hardware replacement, the economics of custom chip design no longer hold up. Consequently, the industry is rapidly moving toward a new model where advanced network software runs on general-purpose, common off-the-shelf hardware sourced from the broader IT industry, a change that fundamentally alters the vendors’ business models and product strategies.
This strategic retreat from custom hardware is now evident across the industry’s key players. Ericsson, long a proponent of its own silicon, is now openly considering a full withdrawal from custom RAN compute development, instead experimenting with virtual RAN solutions that utilize general-purpose CPUs from providers like Intel. The performance gap between custom and general-purpose chips has narrowed significantly, making the switch financially compelling. Similarly, Nokia has made an even more decisive leap by partnering with Nvidia, planning for its 5G Advanced and 6G software to run on the IT giant’s powerful graphics processing units (GPUs). Samsung is also following suit, using equipment such as Dell servers and Intel CPUs in recent network deployments rather than its own dedicated hardware. This coordinated, industry-wide pivot away from proprietary, vertically integrated systems marks a permanent and profound change in the vendor landscape, pushing these companies to become more like software developers than traditional hardware manufacturers.
Software and Virtualization as the New Frontier
This pivot toward general-purpose hardware is a component of a much larger trend: the disaggregation and virtualization of the radio access network. For years, the RAN was a rigid, integrated system where software and hardware from a single vendor were inseparable, creating a strong “vendor lock-in” that operators disliked. The new model breaks this integration, creating a more flexible and open ecosystem where software from one company can run on hardware from another. This shift aligns with the long-held goals of telcos to increase competition among their suppliers, reduce capital expenditures, and gain more granular, software-based control over their network functions. By moving intelligence and processing from dedicated hardware into a cloud-based software layer, operators can deploy new services faster, automate network management, and create a more agile and programmable infrastructure prepared for future demands without requiring constant physical upgrades.
Furthermore, the business case for new 6G hardware is being systematically weakened by rapid advancements in software and artificial intelligence. These new technologies are poised to deliver massive improvements in spectral efficiency, which is the ability to transmit significantly more data over the same amount of licensed spectrum. For instance, recent trials have shown that AI-powered software can produce immediate gains in network performance. Ericsson has already achieved a 10% boost in efficiency through AI-driven link adaptation, while software from startup Cohere Technologies has demonstrated the ability to increase spectral efficiency by up to 50% on a live 5G network. These software-based breakthroughs suggest that much of the anticipated growth in data traffic can be managed by making the existing 5G network smarter and more efficient, further cementing the operators’ position that a costly investment in new 6G radios is simply unnecessary.
A Leaner Industry Forged in Transformation
While this shift toward a software-centric model is strategically necessary, the transition has been accompanied by severe, real-world consequences for the equipment vendors. To adapt to a market defined by continuous software revenue rather than cyclical hardware booms, both Ericsson and Nokia have undergone significant restructuring. Since 2022, the two companies have collectively eliminated over 25,000 jobs as they streamline operations and align their cost structures with the new market reality. Beyond workforce reductions, they are also shrinking their operational scope, with Nokia, for example, exiting non-core markets like cloud infrastructure. Although a contrarian viewpoint exists, with some like U.S. tower company SBA Communications predicting that the relentless growth in wireless data will inevitably force another investment cycle, the overwhelming industry consensus points toward a permanently altered landscape.
The industry that emerged from this period of intense upheaval was fundamentally different from the one that entered it. The decades-old, hardware-defined generational cycle, which had provided a predictable and highly profitable business model, was definitively over, replaced by a more challenging and continuous software-driven paradigm. The major network equipment suppliers were forced to reinvent themselves, shedding their identities as vertically integrated hardware manufacturers to become leaner, more agile, and highly specialized software companies. This painful but necessary transformation ensured their survival, but it also marked the end of an era. The telecom sector that was forged in the aftermath of 5G’s commercial failure became a smaller, more focused industry, forever changed by the realization that the future of mobile connectivity would be written in code, not etched in silicon.
