Qualcomm Stakes Claim to First Purpose-Built 5G Modem for The IoT

Qualcomm’s 315 5G IoT Modem was formally announced on May 21st, 2021. It is targeted at the industrial market, to accelerate cellular IoT adoption and 5G adoption, specifically. The 315 modem is designed to provide gigabit throughput, but with low power consumption and high thermal efficiency. It has a smaller footprint than Qualcomm’s previous 5G chips, with an integrated radio frequency front end (RFFE) and is pin compatible with the manufacturer’s existing 4G Long Term Evolution (LTE) chips. The 315 is being marketed as a global product, offering sub-6 band support in Stand Alone (SA) mode only, and with LTE for fallback. It is designed to allow for as seamless an integration process as possible with existing ethernet and other wired technologies already in use on the factory floor, to eventually displace them in favor of 5G in industrial IoT applications. To this end, Qualcomm is also including extended lifetime software upgrades, and extended hardware maintenance and support, with this new 5G IoT chipset.

Qualcomm claims that the 315 is the first purpose-built 5G modem for the IoT. The manufacturer understandably has a vested interest in accelerating the adoption of 5G for the IoT. Qualcomm was less successful in its 4G intellectual property development than it was for 3G, and this is something it has actively sought to remedy with 5G ever since. The 315 is the first in a line of IoT-specific 5G developments for Qualcomm, with the launch of 7 new 5G Systems on Chip (SoCs) for the IoT coming less than two weeks later. These SoCs do not make use of the 315 modems, however, as they are targeted at other video-centric IoT use cases. Initially stated as sampling at the end of May, and being commercially available in the second half of 2021, IoT module vendor adoption of the Qualcomm 315 5G IoT Modem-RF System has been swift. Fibocom announced its first 315-based modules, the FG130 and FM130 series, immediately prior to Mobile World Congress 2021.

The 315’s features are not merely technical details. Lower power consumption, even for stationary mains-powered deployments, is critical for assisting with thermal efficiency and minimizing heat generation. This matters in industrial environments, where ambient temperatures can exceed 80 degrees centigrade, and be as high as 105 degrees in automotive facilities, to prevent overheating and component failure. A reduced size is also more important than might be imagined for stationary installations, as 5G replacement modules on will need to fit within a specific space, on equipment that currently accommodates wired or Wi-Fi connectivity adaptors. Backwards pin compatibility also simplifies the LTE to 5G upgrade process, encouraging technology migration by minimizing the cost and effort needed to do so. Lastly, longer hardware warranties and support for chipset firmware revisions is vital in the industrial market, as moving to cellular and 5G is a big decision and the equipment will need stay in place for a long time.

The 315 also has some contentious characteristics: it does not support Non-Stand Alone (NSA) 5G networks, does not support millimeter wave (mmWave), and is 3GPP Release 15 compliant only, so it also does not support Ultra Reliable Low Latency (URLLC); the latter long being thought of as important for the industrial IoT. Qualcomm believes that dedicated industrial networks will either use a private network or a slice of a public network, therefore SA is the best option in both cases. The throughput requirements of industrial applications make the higher bitrate of mmWave unnecessary, with the 1.5 Gbps downlink and 330 Mbps uplink of the 315 being purportedly plenty. Finally, Qualcomm asserts that 80% or more of all factory floor connections do not require real-time communications, so can be upgraded now with simpler, cheaper, equipment. Conversely, testing and interoperability for URLLC will take more time, and waiting for that would only unnecessarily delay the industrial 5G IoT market.

There are only a few public 5G SA networks at present, but carrier investment is increasing, and the number will grow. China Mobile, China Telecom, and China Unicom have all reportedly upgraded hundreds of thousands of base stations in total to support SA. While in North America, T-Mobile USA announced the launch of its nationwide 5G SA network in August 2020. In Europe progress seems slower, but at the very end of June 2021, Orange France stated it would launch an experimental 5G SA network in Lannion in July. That is also the pattern that the shipment of Qualcomm’s 315 5G IoT chipset is expected to follow: China first and foremost, followed by North America, and then Europe. This is supported by the initial list of IoT module partners for the 315 as cited by Qualcomm, which are Fibocom, Meig Smart, Quectel, SunseaAIoT (of which SIMCom and Longsung are subsidiaries), and Telit. All but Telit are Chinese vendors, with Telit’s target market for 5G being private networks, for the time being.

That most things 5G-related for the IoT can be done with Release 15 is a mini revelation. 5G take-up in the IoT has been slow, with URLLC being thought of as the one killer feature that may change that. And while URLLC will certainly be relevant to the future, the role of 5G in the medium-term seems much like that of NB-IoT. Namely, to enable applications that were previously unfeasible due to technical, logistical, or economic reasons; but with said technologies not impacting at all on established IoT use cases that can happily continue to be served by LTE. The future link between private network growth and 5G for the IoT seems strong. The wireless factory is Qualcomm’s target market for the 315 in short term, but it plans to push the modem into “retail, energy, automation and manufacturing, precision agricultural, construction, mining, public venues, and more.” Qualcomm is unsurprisingly planning a Release 16 compliant 5G IoT chip in the future too, which will complement the 315, but not replace it.

The achievement of launching an IoT-specific 5G modem should not be understated. It was necessary, and other chipset manufacturers will absolutely be required to do the same in time. Up until now 5G modems intended for consumer mobile broadband (MBB) applications were the only ones available for use in the IoT; typically Qualcomm’s X55 up until now, and its X65 and X62 in the immediate future. But these modems are all too large, too power hungry, overpowered, and over-priced for the IoT in all cases except high end industrial gateways and routers, and Vehicle-to-Everything (V2X) connections. Just as with every other cellular generation, simplified, lower-performance chips, with fewer components, but all the fundamental benefits of the latest standards were required, to create the efficiencies for the required economies of scale. Not all IoT module vendors think a product such as the 315 was needed yet, but it had to happen, and Qualcomm is doubtless proud to have been first.

A limited 100 MHz of bandwidth, so a differently designed RFFE, and no simultaneous transmission, so a single power amplifier, reduces current, power consumption, heat, and size. These are meaningful innovations that also make the 315 50% smaller than earlier 5G modems. Not being too heavily specified and with enough flexibility for the sake of horizontal reusability matters, too. The 315 will be used in ruggedized routers as well as industrial end points, so has Wi-Fi 6 to provide cellular backhaul to multiple attached devices. The 315’s campus-based, private network-centric SA design makes it unsuitable for mobile connectivity, with 5G mobility being more the preserve of NSA NR RedCap modems. But due to the continued dominance of LTE in the IoT and the yet-to-be finalized standards for RedCap, that market is still 3 to 4 years away. In anticipation, Qualcomm wants the 315 to be a template to reduce the time taken for new IoT chipsets to be fielded, as successfully evidenced by Fibocom’s rapid development of new modules.