Renesas' Acquisition of Sequans Validates Specialized Development of IoT Chips

In August 2023, Japanese semiconductor manufacturer Renesas announced that it would purchase French IoT specialist Sequans Communications, for approximately US$249 million. Current expectations are that the acquisition will be finalized by early 2024. Renesas was already active in automotive, industrial, and the Internet of Things (IoT), but not in wide-area connectivity—at least not until recently. In late 2022/early 2023, Renesas launched its first cellular modules, the RYZ014A and RYZ024A, based on the SQN3330 and SQN3430 baseband modems, aka the Monarch and Monarch 2 chips from Sequans. The RYZ014A is Cat-M-only (its Narrowband IoT (NB-IoT) function is disabled in firmware) and the RYZ024A is dual-mode Cat-M and NB-IoT.

The acquisition of a specialist IoT semiconductor manufacturer by a larger, general-purchase semiconductor manufacturer has a recent precedent. Sequans’ purchase immediately brings to mind the acquisition of Altair Semiconductor by Sony Semiconductor, for a comparable US$212 million in 2016, to become Sony Semiconductor Israel. Sequans and Altair started out at about the same time, in 2003 and 2005, respectively, and both sought to innovate in 4G Long-Term Evolution (LTE) specifically, developing not only a new technology, but ultimately a new and parallel market for baseband transceivers, the IoT.

This put them out of direct competition with incumbent manufacturers from the cell phone market, allowing them to develop their  competency on new terms, based on the particular requirements of autonomous machines, namely greater simplicity, less throughput, lower power consumption, and lower cost. With the addition of maximal coverage, this family of features came to be defined as Low Power Wide Area (LPWA), and was formally standardized by The 3rd Generation Partnership Project (3GPP) in the form of LTE-M and NB-IoT. Sequans and its competitor Sony Semiconductor Israel offer more than just LPWA, however. Renesas is doing its due diligence with Sequans in developing the RYZ014A and RYZ024A before making its decision to acquire.

Sequans has seven IoT-specific chips, and only two of them are for LTE-M and NB-IoT. Sequans’ Calliope chips are for Cat-1 and Cat-1bis, its Cassiopeia chips are for Cat-4 and Cat-6, and its most-recent Taurus chips are for 5G New Radio (NR). The IoT is not just about massive Machine-Type Communications (mMTC), as connected devices can be of any size, shape, and value, with any throughput or network technology requirement. The IoT does not exist to only consider a future of millions of connected sensors in a high-density arrangement. It’s just that mMTC stands to exist in the greatest volume, while other cellular connectivity types may represent the greatest value per unit.

In 2022, 40% of all cellular IoT modules shipped worldwide, i.e., the communications sub-systems that contain IoT baseband chips were LTE-M and NB-IoT, while 46% were for Cat-1 and greater. The remainder consisted of the dwindling legacy 2G and 3G, and the nascent and emerging 5G. It makes sense for Renesas to wait to purchase a company that has a mature commercial product offering at just the point that the market for it is starting to ramp up, so that Renesas can now take that to further scale. This is instead of developing something itself in-house for a market that is not yet proven, or reacting once others have already stolen the lead. And although the IoT is not new, the vast majority of its growth potential is still to come.

A recent trend has seen large firms divesting the IoT lines of business that they acquired from specialists, returning them to dedicated providers, having failed to integrate those businesses or to make them scale. Ericsson sold its IoT Accelerator platform to Aeris Communications; Thales shifted its module business into Telit Cinterion; and there have been rumors of Cisco wanting to sell its IoT Control Center and of wireless carriers planning to sell off their IoT business units. But semiconductor manufacturers should not risk falling afoul of this, as they do not simply sell products and services, but own legally-protected Intellectual Property (IP) that can be evolved into other products or licensed.

In acquiring Sequans, Renesas is buying into cellular LPWA just as Cat-M starts to become a significant force, with 70 million Cat-M modules shipping in 2022. Renesas will also reap the benefits of Sequans’ knowledge of optimizing broadband IoT chips. The work Sequans has done with 5G may prove foundational for the best possible chance of success in 5G NR Reduced Capability (RedCap) should Renesas choose to take that route. RedCap Release 18 is hoped to be a big enabler for the IoT, and Renesas could combine Sequans’ baseband competence with its own ancillary component expertise in processors, memory, power management, amplifiers, and sensors to create compelling and highly-integrated RedCap modules or Systems in Package (SiPs).

Sequans seems likely to continue on as a specialist subsidiary of Renesas, just as Altair does now for Sony. Sequans in its new guise will likely receive a cash injection, and access to other core competencies from its new parent that will accelerate product development, especially on the software side potentially. Software excellence is becoming very important in the IoT chipset market, for maximal ease of application development and physical integration into IoT systems. The IoT chipset champions of the future will be chosen on the strength of their software first, especially if licensing-based models come to replace the present day payment for units of IoT chips and module hardware.

There are still more IoT chipset specialists that could become hot acquisition targets for larger, general-purpose manufacturers, notably Nordic Semiconductor. The first wave of Cat-M and NB-IoT chipsets came from big cell phone semiconductor manufacturers, which raced to market excited by visions of massive sales volumes. But their enthusiasm either waned with time or died entirely. Qualcomm switched from internal development to third-party licensing, MediaTek exited the market, and HiSilicon (Huawei) was forced out by a lack of international acceptance and sanctions. But if Qualcomm or MediaTek ever want a second shot, or if any of the semiconductor manufacturers that were big in 2G and 3G ever want to get back into cellular, this may be where to look.