Qualcomm’s First NB-IoT-Only Chipset Expands and Focuses Its IoT Strategy

Subscribe To Download This Insight

By Jamie Moss | 2Q 2020 | IN-5823

Qualcomm’s new cellular Internet of Things (IoT) chipset continues its legacy of functional differentiation through silicon integration, at the same time as showing a new level of commitment to the Narrow Band IoT (NB-IoT) market specifically. Industry players must pay attention to a new shift toward customizable, application-specific IoT chipset development.

Registered users can unlock up to five pieces of premium content each month.

Log in or register to unlock this Insight.

 

Integrate to Simplify

NEWS


Qualcomm is launching its first Narrowband Internet of Things (NB-IoT)-only chipset, the 212 LTE IoT modem, and claims it will be the most power-efficient cellular Low Power Wide Area (LPWA) chipset seen to date. The 212 LTE IoT modem integrates onto a single die the NB-IoT baseband processor, Random Access Memory (RAM), Power Manager (PM), Power Amplifier (PA), and Radio Frequency Front End (RFFE) transceiver. It is being pushed as effectively a System on Chip (SOC) product. With a purported total of 32 components, compared to the claimed 120 to 150 components used by competing chipsets, Qualcomm is cutting complexity to maximally reduce cost, size, and power consumption. The 212 LTE IoT modem will measure 10 mm square and will begin sampling in the second half of 2020.

A key feature of the new 212 LTE IoT modem is that its baseband processor uses a Cortex M3. Its peak clock speed is intentionally limited to a mere 200 MHz for the sake of cost but, innovatively, 80 MHz of headroom is exclusively budgeted for running third-party Original Equipment Manufacturer (OEM) applications, to tailor the chipset to perform for the benefit of discreet IoT use cases. Flash memory is used to execute code in place on the Cortex M3, and a Software Development Kit (SDK) with Application Programming Interfaces (APIs) will be provided to program it. By contrast, the Cortex M0 used in Hisilicon’s Hi2115, the single-mode NB-IoT modem market leader by volume, cannot run a protocol stack or applications. It requires an external processor, ergo more cost.

Qualcomm’s 212 LTE IoT modem supports NB2 standards, and global frequency bands from 700 MHz to 2.1 GHz, incorporating Observed Time Difference of Arrival (OTDOA) for embedded cellular positioning. 0.8 microamps is being claimed for its deep sleep mode, a massive 70% lower than Qualcomm’s MDM9205 LPWA chipset from mid-2019, and it is designed to operate with as low as 2.1 volts so that even aging batteries can power it for a long IoT device lifespan. The 212 LTE IoT modem will be targeted at applications like but not limited to metering, where it sleeps most of the time, periodically but infrequently waking to monitor a communications channel. Three module vendors are working on designs using the chip, and it will be certified with a Tier-1 U.S. carrier in September.

Last Man Standing

IMPACT


Qualcomm was quick to enter the cellular LPWA market with its MDM9207, MDM9206, and MDM9205 chipsets. The MDM9206 and MDM9205 have proved popular with module vendors worldwide. But while major competitor HiSilicon has only ever sold single-mode NB-IoT chipsets, Qualcomm has been committed to multi-mode. Both the MDM9206 and MDM 9205 are LTE-M and NB-IoT chipsets, with 2G fallback. This has allowed the semiconductor vendor to hedge its bets and use the same product to supply customers who want either LTE-M or NB-IoT as well as providing future proofing for customers who currently require only 2G. Qualcomm has stood alone among its competition in not developing a single-mode NB-IoT design.

Multi-mode chipsets allow for single Stock Keeping Unit (SKU) IoT device deployments that make use of the right technology with the best coverage, at any location in any country. This is convenient when LTE-M and NB-IoT are still being rolled out and international coverage is patchy. LPWA-based IoT device vendors want to minimize variants, to cut development costs, for affordability savings that can be passed on to their customers along with guarantees of continuity of service availability, which is a critical requirement in the enterprise sector. It is the lack of ubiquity of LPWA networks that has driven the Cat-1 market to unexpected levels of success as a convenient general-purpose intermediary, Qualcomm’s first dedicated IoT chipset, the MDM9207, being Cat-1.

However, multi-mode chipsets are not good for the long-term mass market affordability of NB-IoT. Building in LTE-M and 2G doubles the cost, preventing NB-IoT prices from falling below US$3 per unit while maintaining an acceptable profit margin. 2G fallback also negatively impacts power efficiency. The 212 LTE IoT modem is a calculated shift in strategy for Qualcomm, from integration in breadth to integration in depth, to focus on the intensity of utility of a single radio standard rather than the inclusive integration of as many radio standards as possible. Qualcomm is committing itself to the potential of the NB-IoT market specifically, and for the first time will be competing directly with HiSilicon, no longer restricting its growth in the Chinese market and upping the ante through the embedding of application-specific customization.

Next-Level Challenge

RECOMMENDATIONS


Reconciling between the low power raison d’être of cellular LPWA and the high power consumption of Global Navigation Satellite System (GNSS)-based positioning is a problem for NB-IoT. For optimal power consumption in battery-powered devices, NB-IoT is best suited to making infrequent connections and occasional location reporting rather than Real Time Location Services (RTLS). Of the 30-plus NB-IoT chipsets launched only five have featured integrated GNSS, none being single-mode NB-IoT. But the track and trace application opportunities are too numerous to ignore and have even greater need of minimal cost, chipset-based devices than traditional NB-IoT markets like water metering. To this end, Qualcomm’s 212 LTE IoT modem’s support of OTDOA serves up cellular triangulation as a GNSS replacement, giving a further clue to its long-term strategy.

OTDOA uses the coordinate location of cellular base stations as reference points for calculating the position of a mobile terminal, based on the relative time delay in the reception of its transmissions. Wireless carriers do not share their cell site maps. So OTDOA devices are only useful for resolving a location within the confines of the home carrier’s network. With the 212 LTE IoT modem, Qualcomm may be looking to serve the national, carrier-specific IoT device market. While its long-term utility is secured through its future as part of the 5G standards, NB-IoT needs a boost outside of China. Carriers have struggled with the chicken-and-egg situation of the opportunity of NB-IoT enterprise demand versus the cost of NB-IoT coverage rollout, compounded by an adherence to legacy pricing models based on the incremental consumption of data.

NB-IoT industry participants must therefore be alert to a shift from a general-purpose to an application-specific, and even network-specific, cellular LPWA chipset development paradigm courtesy of the proliferation of software-defined frequency band tuning and application-based Microcontroller Unit (MCU) customization. Horizontal developments are generally to be encouraged, as they offer maximum re-salability, to hit as many different use cases as possible, for maximum scale and profitability. But vertical development is absolutely to be supported whenever a single market opportunity is valuable enough—the classic example in the IoT being automotive. The 212 LTE IoT modem looks a lot like Qualcomm’s NB-IoT spearhead to this next-level challenge.

 

Services