The ABI Research team has been traveling for the past few months, meeting with clients (including Ericsson, Nokia, NVIDIA, and Intel), and having deep discussions on the state of the telecommunications industry and its future. This blog summarizes a few of these findings/trends and how we expect the market to evolve going forward.
Silicon Superiority for 5G-Advanced and 6G
In the early days of 5G, Huawei gained a major strategic advantage by investing proactively in Gallium Nitride (GaN) Power Amplifier (PA) silicon, and hence the ability to manufacture energy-efficient and powerful Massive Multiple Input, Multiple Output (mMIMO) radios. Eventually, the market caught up, and all vendors now have access to GaN PAs.
But as of 2025, the tide has reversed for Ericsson, Nokia, and Samsung: they have access to foundries for cutting-edge semiconductor manufacturing—Intel 18A (1.8 Nanometers (nm)), Samsung SF2 (2 nm), and TSMC N2 (2 nm). In contrast, Huawei only has access to SMIC and other local foundries, which are still on 7 nm, without an immediate roadmap to newer processes.
Radio processing is limited in terms of Total Cost of Ownership (TCO) and thermal load, meaning that 18A will provide a significant advantage to Western vendors, as they manufacture faster chips with higher energy efficiency. The very same advantages will carry over to 6G, where Artificial Intelligence (AI) inference capabilities can be integrated into these processors, either integrated or through AI chiplets. The race for 6G is just starting, but silicon superiority may be a critical factor for vendor dominance going forward.
All Eyes on ISAC
Interest in Integrated Sensing and Communication (ISAC) is intensifying, which is a new use case for 6G. In fact, it is one of the areas on which most parties from The 3rd Generation Partnership Project (3GPP) agree that it should be a priority for the new generation and has the capability to really transform mobile network deployments and business models.
As of the end of 2025, the telecom industry is working to integrate sensing capabilities into 5G-Advanced and is planning to do so for 6G, using the same waveforms and Orthogonal Frequency Division Multiplexing (OFDM) air interface. There is still monumental effort needed to make ISAC work and even more to commercialize it in a public environment, but 3GPP has shown in the past that it excels in these scenarios.
Initial use cases illustrate significant potential, but there is work to be done to advertise these capabilities to enterprises; according to informal discussions ABI Research has had recently, the Chinese pre-standard commercial trials for ISAC (based on 5G-Advanced) have not yet produced tangible revenue. This is somewhat expected, as it is quite early in the technological development process and enterprises may not be willing to adopt a very new, and unproven, technology.
Nevertheless, many telcos are now starting to pivot from initial 6G reluctance to curiosity and trying to understand how they can monetize ISAC. As a result, expect the next steps described below.
6G Hype Cycle Begins
All indications point to the beginning of the 6G hype cycle and ABI Research expects that MWC 2026 will be partly devoted to early 6G discussions and achievements. Broader market conditions are converging toward the advanced capabilities of 6G, including the mass market deployment of Standalone (SA), availability of advanced network capabilities (like Quality of Experience (QoE) Application Programming Interfaces (APIs)), and the ongoing pivot of telcos toward AI factories and capabilities. ISAC is also a main vehicle for these innovations and is expected to become a key centerpiece for the new generation.
The NVIDIA-Nokia partnership and NVIDIA’s relentless growth have accelerated these discussions even further and is bringing 6G and AI/Generative AI (Gen AI) closer.
Telcos have the unique opportunity to capitalize on their assets for 6G: they have the real estate (central offices, distribution centers, and cell sites) and the transport network. NVIDIA is now accelerating the merging of 6G and AI, putting telcos in the unique position of having distributed connected computing assets. They need to take this opportunity seriously and use it to diversify their business model.
DAS/DRS Is Merging with Private Cellular
In-building wireless systems are starting to merge with private cellular, or offer the capability for various types of private cellular capabilities. With support for Multi-Operator Core Networks (MOCN), multiple frequencies, shared spectrum, and more, modern in-building wireless systems like Ericsson Dot or Huawei Lampsite are in an excellent position to connect the enterprise, but also offer communication limited to on-premises systems only.
This is creating an opportunity for 5G, where it offers distinct advantages over Wi-Fi, especially as laptops with 5G modems and embedded Subscriber Identity Modules (eSIMs) are making a comeback. 5G systems offer better coverage, support for mobility, no need to connect to insecure Wi-Fi networks (e.g., in airports or hotels), and better remote management capabilities. For small enterprises, the public network can be used to provide connectivity for small workforces without the need to deploy on-premises networks.
In addition to the above benefits, Distributed Antenna Systems (DASs)/Distributed Radio Systems (DRSs) may be a gateway for edge AI capabilities. For example, a DRS deployment typically includes the radios (multiple on a floor), a radio unit (one per floor) and a baseband processor, typically deployed in a cabinet in the building computing space.
This baseband processor may become a key point for deploying edge AI capabilities, creating an opportunity for infrastructure vendors and operators that are already controlling these in-building wireless relationships.
Custom Silicon Versus COTS for 6G
It is not yet clear what type of silicon will dominate 6G Radio Access Networks (RANs), but general market consensus is that it should be diversified across custom silicon, Graphics Processing Units (GPUs), and Central Processing Units (CPUs). Will large infrastructure vendors be successful in porting their baseband software to CPUs and GPUs? So far, custom silicon has delivered an order of magnitude lower performance and energy efficiency compared to Commercial-off-the-Shelf (COTS) solutions, but this may change with NVIDIA seeding the market.
Nokia aims to port its baseband software to NVIDIA’s platform and conduct early field trials in 2026, opening the way for more advanced use cases and trials. But longer term, ABI Research expects custom silicon to still play a large role in 6G RAN, especially for mMIMO, where silicon is highly optimized. It is also very likely that operators will soon intensify their focus on their urban networks where 6G and AI will likely be deployed, due to advances in satellite technology.
5G NTN Will Rewrite the Rules of Rural Coverage
Direct-to-Cellular (D2C) satellite systems are entering commercial use, and several mobile operators have struck deals with satellite providers. Although current Low Earth Orbit (LEO) satellite constellations are not fully 3GPP-compliant, we expect next year to see many Non-Terrestrial Network (NTN)-compliant devices in the market, but some constellations and LEO systems will remain on proprietary protocols to maintain high performance. Some of them are even aiming to become the global cloud backbone. Regardless of NTN compatibility, satellite integration with cellular networks will continue, either through standardized NTN partnerships or proprietary roaming agreements (like the T-Mobile US – Starlink partnership).
The result is that ubiquitous, but simple, coverage will be available nationwide and in rural or very remote and isolated areas, assuming a user has access to open sky. This may eventually lead to relaxed terrestrial coverage obligations in rural areas, where satellites can provide voice, messaging, and, most importantly, emergency communications. Maintaining rural cellular networks takes up a significant part of telco financials, while most—if not all—of these sites generate a modicum of revenue compared to sites in more populated areas.
Satellite can indeed reduce rural network costs, and can help operators focus more on urban areas, where they are generating most of their revenue. More focus and effort in these areas means that operators will now have more operational and financial capacity to take advantage of the AI opportunity.
For further insight into the future of telecom, read ABI Insights across our 5G, 6G, & Open RAN, Enterprise Connectivity, Telco AI, and Space Technologies & Innovation research services.