By: Ohad Peled, Product Marketing Manager, Sony Semiconductor Israel
Compute power gravitates towards the edge
As IoT ecosystems evolve, end-to-end cloud solutions designed to generate insights based on the analysis of IoT devices’ ambient data are being created and deployed. Processing data at the network edge in devices that have the requisite computing power enables autonomous valuable operation, even when offline, by generating local insights that can trigger immediate actions. Traditionally, we tend associate increased compute power with increased power consumption. However, in some cases, boosting the device’s compute power leads to significant power savings in end-to-end IoT solutions.
In most wireless IoT devices, network connectivity consumes the major share of the power. Even for the most efficient and optimized devices, powering up the modem and the RF chain to send and receive data from the network is very expensive in terms of energy consumption. Therefore, when devices can intelligently manage communication interactions and limit unnecessary connections, they save precious power over time – resulting in longer device lifetimes.
Lower bandwidth connectivity
In typical applications, devices gather data from their environment and transmit it back to the cloud for analysis. There is a growing need to control the amount of this data. The ability to process data at the edge helps minimize network communication by sending only pertinent data to the server. It also improves cloud analysis performance and accuracy. Devices can use their integrated processor to continuously monitor collected parameters and use network connectivity less frequently. For example, employing it for regular status updates or when a preset threshold is crossed, thereby reducing network traffic and saving power. Another, more complicated application example, could be a device that collects continuous data from a machine for preventive maintenance purposes, analyzing it using dedicated AI algorithms and updating the cloud only about the state of the machine or when unscheduled maintenance could be beneficial.
Reducing the amount of data that is transmitted also allows devices to use network protocols that are specifically designed for low power and low data rate applications, namely LTE-M and NB-IoT. These protocols allow the device to stay connected to the network when the power-hungry modem and RF chains are not communicating for extended periods of time. Less network traffic also translates directly to cost savings on data usage, which enhances the business case and the margins.
Putting it all Together
Obviously, the benefits outlined above come with a tradeoff. In addition to the rigorous security and device management capabilities, which are fundamental to any IoT device, devices now need to allocate precious compute power to analyzing, screening, and compressing the data it aggregates. This drives more integrated solutions across the supply chain down to the IoT chipset level.
Sony’s Altair Edge Offering
Sony’s Altair IoT chipsets are built from the ground up to support edge device requirements, enabling much of the data processing to be executed on the device side. Their dedicated core for application development is detached from modem operations. This allows the application to take full advantage of a robust Arm Cortex-M core, including AI engine capabilities, which facilitate real-time analysis, and corresponding actions on the device side. In turn, this reduces the amount of data needed for communications over the air and improves the accuracy of the data analysis process that is conducted in the cloud.
The other major advantage of Sony’s Altair ALT1250 and ALT1255 cellular IoT chipsets is the security architecture. Security in the IoT era is crucial. The company’s chipsets employ a multi-layered security architecture that enables the highest level of security for applications running on the chipset. It features chip and device hacking-protection as well as a robust service access layer. Sony’s Altair chipsets provide the most advanced hardware-based tools to further ensure end-to-end security. This gives our customers the ability to develop applications on the device side without worrying about the majority of security threats, as they are secured by design.