ST combines sensors and edge processor for new ISPU concept - Embedded.com

ST combines sensors and edge processor for new ISPU concept

The ISPU takes edge intelligence one step further to combine signal processing and AI algorithms onto MEMS sensors, injecting local decision-making and saving space over system-in-package devices.

Advertisement

STMicroelectronics is combining sensor and digital signal processor (DSP) onto the same silicon to enable the concept of the intelligent sensor processing unit, or ISPU. The company said this integrates brains into sensors to facilitate an “onlife era” enabled by smart sensors that are able to sense, process, and take actions.

Many companies talk about enabling localized, intelligent decision making at the edge to overcome many issues such as latency, security, privacy and bandwidth availability. But with the ISPU, STMicroelectronics takes the edge capability one step further to combine signal processing and AI algorithms onto MEMS sensors, injecting local decision-making and further saving space over system-in-package devices, and cutting power by up to 80% – ST’s calculations show a 5-6x saving over system-in-package approaches in sensor-fusion applications; they also show a 2-3x saving in RUN mode.

STMicroelectronics ISPU concept
ST said it is enabling the migration of intelligent processing from ‘on the edge’ to ‘in the edge’. (Source: STMicroelectronics)

In its announcement, ST said its ISPU provides substantial benefits in power consumption, packaging, performance, and price. The proprietary ultra-low-power DSP can be programmed in C, familiar to many engineers. It also allows quantized AI sensors to support full- to single-bit-precision neural networks. This ensures good accuracy and efficiency in tasks such as activity recognition and anomaly detection by analyzing inertial data.

ST’s proprietary, C-language-programmable DSP is an enhanced 32-bit reduced instruction set computing (RISC) machine. It is extensible (in the chip-design phase) for dedicated instructions and hardware components. The processor offers a full-precision floating-point unit, uses a fast four-stage pipeline, operates from 16-bit variable-length instructions, and includes a single-cycle 16-bit multiplier. Interrupt response is a spritely four cycles. ST’s sensors with ISPUs will be packaged in standard 3mm x 2.5mm x 0.83mm packages and will be pin compatible with their (ST) predecessors, allowing quick upgrades.

Commenting on the significance of the ISPU, Andrea Onetti, executive vice president, MEMS sub-group for STMicroelectronics, said, “While technically challenging, integrating ST’s sensors on the same piece of silicon with our ISPU does improve sensor-based systems from an online experience to an onlife one. It advances the sensor’s features to speed decision-making by reducing data transfers, enhancing privacy by keeping data local, while reducing size and power consumption, which cuts costs. Moreover, the ISPU is easily programmable with commercial AI models and can ultimately operate with all of the leading AI tools.” The company said that with some 10 million plus C language developers, its new ISPU will enable many to deploy this intelligence capability at the edge.

In its launch material, ST talks about a shift in society to an ‘onlife era’ which necessitates this level of integration that its new ISPU offers. It states, “The onlife era acknowledges living with continuous assistance from connected technologies, enjoying natural, transparent interactions, and seamless transitions, with no discernible distinction between online and offline. With the ISPU, ST is enabling this era by helping to migrate intelligent processing into sensors that support the fabric of life: no longer at the edge but in the edge.”

First devices in the family of ISPUs are expected in Q2 2022.


Related Content:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.