Aspinity Joins STMicroelectronics Partner Program to Deliver Complete AnalogML Solutions for Always-on Edge Processing

Analog machine learning pioneer aligns with leader in high-performance microcontrollers to support extended battery life in IoT devices


PITTSBURGH, March 02, 2021 (GLOBE NEWSWIRE) -- Aspinity has joined the ST Partner Program to accelerate the integration of Aspinity’s ultra-low-power analog machine learning (analogML™) core with STMicroelectronics’ high-performance microcontrollers (MCUs) into small, power-constrained always-on sensing devices.

Aspinity’s products are based on its near-zero-power analogML core, which enables event-based system wake-up to occur completely within the analog domain at the front end of the signal chain—keeping the majority of the system asleep until an event is detected. The collaboration between Aspinity and STMicroelectronics enables customers to rapidly adopt analogML to bridge the gap more efficiently between the analog physical world and the digital processing world. The companies achieve this by focusing data digitization and higher-order analysis only on important data.

Aspinity has already released the first platform based on its collaboration with ST: its Voice-First Evaluation Kit (EVK2), a complete hardware/software development kit that demonstrates a full end-to-end, ultra-low-power analog voice wake-up and preroll solution using the STM32H743ZI MCU. Aspinity’s analogML core uses less than 25µA at the front end of the voice-first system to detect voice from raw analog microphone data while simultaneously and continuously collecting and compressing the 500ms of analog preroll data—which is what’s required by most WWEs to accurately determine that a command has been spoken. When voice is detected, the analogML core in the EVK2 wakes the STM32H743ZI to deliver the preroll for accurate keyword analysis by the included Amazon WWE. In voice-first applications where voice is present only 10-20% of the time, such as in voice-enabled TV remotes, this low-power, high-performance architectural approach extends battery life by up to 10x.

“ST is a global leader in high-performance, low-power microcontrollers and microprocessors for mass-market applications. Together with Aspinity, we are expanding the toolbox for customers that seek to find the perfect balance between power consumption, features, and system performance in always-on sensing applications,” said Daniel Colonna, marketing director, MCU Division, STMicroelectronics. “This collaboration speeds time-to-market for an exciting new way to extend battery life in always-on sensing devices.”

“Aspinity and ST technologies are highly complementary, and becoming an ST Authorized Partner allows each company to bring together its expertise to deliver to our customers high performance in an extremely power-efficient way,” said Tom Doyle, CEO and founder, Aspinity, Inc. “While we’ve focused primarily on voice-first applications to date, the STM32 series of MCUs, together with our programmable analogML core, offers a robust and flexible combination that can support a wide range of always-on sensing applications. These include acoustic event detection, like glass break and alarms, as well as industrial vibration monitoring. We look forward to continuing to develop innovative solutions together.”

ST Partner Program

STMicroelectronics, a global semiconductor leader serving customers across the spectrum of electronics applications, created the ST Partner Program to speed customer development efforts by identifying and highlighting companies with complementary products and services. Moreover, the program’s certification process assures that all partners are periodically vetted for quality and competence. For more information, please visit www.st.com/partners

About Aspinity

Aspinity is the world leader in the design and development of analog processing chips that are revolutionizing the power- and data-efficiency of always-on sensing architectures. By delivering highly discriminating analog event detection, Aspinity’s ultra-low power, trainable and programmable analog machine learning (analogML) core eliminates the power penalty of moving irrelevant data through the digital processing system, dramatically improving battery life in consumer, IoT, industrial and medical applications.

To learn more about Aspinity, visit www.Aspinity.com. On Twitter: @aspinity.

A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/04b90e3c-c105-4a76-8fec-bebd22210edd

 
Aspinity’s Voice-First Evaluation Kit includes ST’s STM32H743ZI MCU

Coordonnées