Sony Semiconductor Israel to Return to Independent Operation as Altair Semiconductor

During 1H 2026, it is expected that Sony Corporation in Japan will spin out the Sony Semiconductor Solutions subsidiary Sony Semiconductor Israel. Sony Semiconductor Israel will continue operating independently under its original company name of Altair Semiconductor, and Sony Corporation will continue to be an important investor, if no longer the majority shareholder. Altair Semiconductor is an Internet of Things (IoT) Low-Power Wide Area (LPWA) specialist fabless chipset manufacturer that was acquired by Sony in 2016 for US$212 million. As this is not expected to be an outright sale, no formal valuation of Sony Semiconductor Israel has been prepared by its parent; however, in 2025, the Altair subsidiary shipped in excess of 12 million LPWA chipsets, with a  strong sales pipeline that suggests even higher numbers in 2026.

In 2016, at the time of Altair’s acquisition, cellular IoT was ramping up substantially in expectation, following the standardization of the LTE Cat-M (Machine-Type) and Narrowband-IoT (NB-IoT) air interfaces in The 3rd Generation Partnership Project (3GPP) Release 13. Cat-M and NB-IoT were the first IoT-specific cellular technologies, with the world’s first two Cat-M networks being launched in South Korea in 2016, alongside three NB-IoT network launches in Germany, the Netherlands, and Norway. Altair Semiconductor cut its teeth on LTE Cat-1 before specializing in Cat-M and NB-IoT. Altair has openly rejected the development of LTE Cat-1bis chipsets stating that the window of opportunity is too small time-wise for Western players, as LTE will inevitably be replaced by 5G with carriers seeking to re-farm their 4G spectrum; meanwhile, Cat-M will not be rendered obsolete, having been rolled into the 5G standard for low-power, low-cost use cases.

The promise of Cat-M took longer to become a reality than all invested parties—wireless carriers, semiconductor manufacturers, and IoT module vendors—anticipated. It was only with the eventual sunset of 2G networks that Cat-M demand blossomed, with 2021 and 2022 being boom years, some 5 years after the industry hype around the technology began. In 2020, Cat-M represented 4% of cellular module shipments for the IoT, which increased to an anticipated 17% by the end of 2025, with a cumulative total of 354 million Cat-M modules being shipped since 2020. Altair Semiconductor’s chipsets are roughly 20% of the Cat-M market, and as enablers of massive IoT connectivity, Cat-M chipsets are necessarily low cost, with modest margins. But Altair’s ALT1350 is renowned by IoT device manufacturers and semiconductor competitors as among the most capable and advanced.

Perhaps if the Cat-M market had taken off sooner, or had been worth more, Sony Corporation’s decision could have been different. Despite the quantities of Cat-M modules shipped, the technology only accounted for 9% of total IoT module revenue in 2025. Additionally NB-IoT still struggles to find viable applications outside of China, where there is no Cat-M, and even within China, NB-IoT is now being replaced by LTE Cat-1bis. The industry has also reached a point where few additional Cat-M networks will be launched. NB-IoT was designed for stationary IoT devices, and so was more universally applicable, even to smaller domestic carriers, achieving a greater number of network deployments than Cat-M, with ABI Research counting 120 live networks at the end of 2025. Cat-M was designed for mobile IoT devices, so it was of most interest to larger carriers that receive and are responsible for more roaming traffic, leading to 91 networks being deployed by the end of 2025.

But Cat-M is specified to operate in specific LTE frequency bands only, so this requires carriers to set aside some spectrum resources for as long as they want to maintain a Cat-M network. And as long as carriers host devices that use LTE-M, they will be obligated to do so by their customers. Now that we are halfway into the 5G era, LTE refarming is foremost in carriers’ minds, as is the broadband consumer market that constitutes the majority of their business. If carriers have not seen sufficient value in launching Cat-M yet, they are unlikely to now that LTE spectrum reuse is so pressing. In recent years, Sony Corporation has focused on entertainment, gaming, and content; as well as image sensors specifically in regard to semiconductors. Meanwhile, to enlarge its addressable market, Altair has announced the ALT1550 Enhanced Reduced Capability (eRedCap) chip to target consumer AI and IoT devices. All of this is not to say that the Cat-M and cellular market is not worthwhile, but simply that the potential, and the effort required to achieve it, is no longer in alignment with Sony’s goals.

The spin-out and reestablishment of Altair Semiconductor is part of a wider trend from across the entirety of the IoT of large companies that acquired mature IoT specialists with a view to making them scale, now divesting said acquisitions. This allows the parent companies to refocus their efforts on their core competency and primary sources of revenue during more challenging financial times, having potentially overextended their interests into too many lines of business. It also returns the specialist entities to more expert management with simplified goals, no longer having to compete internally for budget with larger lines of business, or needing to be strategically allied or consciously synergistic to those other businesses’ products.

In many cases, the parent companies have not been cutting their one-time subsidiaries adrift, but maintain a controlling financial interest—at least initially—while inviting other parties with greater alignment to the ex-subsidiary to purchase shares, to more organically grow the divested business through aligned business expertise. Recent examples, in addition to Altair Semiconductor’s separation from Sony Corporation, was the sale of Nokia’s IoT Device Management Platform division to Lumine Group in 2023, the sale of Ericsson’s IoT Accelerator Connectivity Management Platform (CMP) to Aeris also in 2023, and the carve-out of Vodafone’s internally incubated Vodafone IoT Mobile Virtual Network Operator (MVNO) connectivity business in 2024.

Cat-M is not expected to be replaced formally in 3GPP standards until the advent of 6G. And while the range of IoT applications that Cat-M is suited to is unlikely to expand much further, Altair is free to develop chipsets for future cellular IoT technologies; for example, eRedCap as announced at CES 2026, as well as connectivity-adjacent IoT compute technologies. Sony Semiconductor’s Research and Development (R&D) center in Hod Hasharon had a round of layoffs in mid-2025 in preparation for the later divestiture. It was a bitter pill for Altair to swallow, but was one of several Sony divisions to experience such a slim down, and Altair’s Intellectual Property (IP) and engineering competency was never in question. A buyer has yet to be found, and Sony Corporation is still glad to financially back Altair for the time being even if in a reduced capacity.