first demo of Arm Cortex-M85, automotive IMU with ML

This week’s roundup of embedded industry news and products below features news from: Bosch Sensortec, Cincoze, Dispelix, Evonik, Infineon Technologies, InnovationLab, Nvidia, Qeexo, Renesas Electronics, SiMa.ai, STMicroelectronics, TriLite, u-blox.

SiMa.ai has appointed Lip-Bu Tan to its board of directors and announced the closing of an additional $30 million funding, closing its series B round and bringing the total money raised to date to $150 million. Tan, as former CEO of Cadence Design Systems, and chairman of venture capital firm Walden International said, “Machine learning has had a profound impact on the cloud and mobile markets over the past decade and the next battleground is the multi-trillion-dollar embedded edge market. SiMa.ai has created a software-centric, purpose-built MLSoC platform that exclusively targets this large market opportunity.”

Renesas Electronics is to present the first live demonstration of a MCU based on the recently announced Arm Cortex-M85 processor at Embedded World 2022 in Nuremburg, Germany from June 21-23. The Arm Cortex-M85 processor features Helium technology, Arm’s M-Profile Vector Extension that enables advanced DSP/ML capabilities and helps accelerate compute intensive applications such as endpoint AI. Renesas is developing a new series of RA MCUs based on the Cortex-M85 processor, planned for release in 2023. The new RA devices will bridge the gap between MCUs and MPUs, enabling complex and compute-intensive applications with the lower power consumption and ease of use of an MCU.

STMicroelectronics has released the ASM330LHHX inertial measurement unit (IMU), which it said is the first automotive IMU with an embedded machine learning (ML) core. The ML core enables fast real-time response and sophisticated functions with low system power demand. The automotive-qualified IMU leverages ST’s MEMS technology, housing a 3-axis accelerometer and 3-axis gyroscope in a 2.5mm x 3mm x 0.83mm outline. The 6-axis module provides movement and attitude sensing for functions including vehicle positioning and digital stabilization. The ML core, a hardwired processing engine, runs AI algorithms directly on the sensor, ensuring extremely low latency between sensing an event and the vehicle’s response.

Infineon Technologies said it is bringing its AIROC Wi-Fi and Bluetooth connectivity solutions to the Nvidia Jetson edge AI platform. AI-at-the-edge devices require reliable and instant Wi-Fi cloud connectivity for lifecycle management functions like deploying AI models through over-the-air (OTA) updates. Infineon’s AIROC Wi-Fi and Bluetooth combo portfolio provides high-performance Wi-Fi 6/6E solutions and low-power Wi-Fi 5 solutions that are integrated with the full range of Nvidia Jetson system-on-modules. AI-enabled edge devices will benefit from robust connection, power saving, low latency, extended range and seamless data streaming with Infineon AIROC products.

Nvidia announced that Taiwan’s leading computer makers are set to release the first wave of systems powered by the Nvidia Grace CPU superchip and Grace Hopper superchip for a wide range of workloads spanning digital twins, AI, high performance computing, cloud graphics and gaming. Dozens of server models from ASUS, Foxconn Industrial Internet, GIGABYTE, QCT, Supermicro and Wiwynn are expected starting in the first half of 2023. The Grace-powered systems will join x86 and other Arm-based servers to offer customers a broad range of choice for achieving high performance and efficiency in their data centers.

Cincoze has introduced the DV-1000, a high-performance embedded computer with a compact chassis and flexible expansion options, as part of its DIAMOND product line. Targeting industrial applications with limited installation space, such as smart manufacturing, machine vision, and railway computing it has a footprint of only half a sheet of A4 paper, with a chassis measuring 224 × 162 × 64 mm and is powered by a 9/8th Gen Intel Core i7/i5/i3 CPU (Coffee Lake-R S series) that supports up to 128 GB of DDR4 2666 MHz memory. Storage options include 1x 2.5 SATA HDD/SSD tray, 2x mSATA slots, and 1x M.2 Key M 2280 slot for a high-speed NVMe SSD.

u-blox has announced the u-blox LARA-L6 LTE Cat 4 cellular module. The smallest of its kind to offer global coverage and 2G/3G fallback, LARA-L6 combines high data throughput and native support for external u blox GNSS receivers with a full set of security features and delivers download data rates of 150 Mbps and upload data rates of 50 Mbps. Measuring only 24 x 26 x 2.6 mm, it is available in two variants: a global one, the LARA-L6004, and a North American one, the LARA-L6404. It comes with the most relevant regulatory and MNO certifications, which greatly simplifies logistic management for worldwide deployment, and supports a comprehensive set of security features including a root of trust, secure boot, and secure updates.

Qeexo and Bosch Sensortec announced that machine learning algorithms created using Qeexo’s AutoML can now be deployed on Arduino Nicla Sense ME with Bosch BHI260AP and BME688 sensors. Qeexo AutoML is an automated machine-learning (ML) platform that accelerates the development of tinyML models for the edge.  Bosch’s BHI260AP self-learning AI sensor with integrated IMU, and BME688, a 4-in-1 gas sensor with AI, significantly reduce overall system power consumption while supporting a wide range of applications for different segments of the IoT market.

TriLite announced it is working with Dispelix, a provider of see-through waveguides for wearables, to create an ultra-compact display system for augmented reality (AR) smart glasses. The system combines TriLite’s Trixel 3 laser beam scanner (LBS) with Dispelix’s advanced waveguides. For AR glasses, the image creation system requires a projector, such as Trixel 3, as well as a waveguide, which is a predominately transparent, thin piece of glass or plastic that bends and combines light beams from multiple sources into the eye. The company said the combined system form factor with Dispelix’s ultra-thin waveguide is one of the most compact ever achieved – meaning that AR will be possible in every pair of glasses, regardless of size or style.

InnovationLab has acquired the TAeTTOOz printable battery technology from Evonik, a technology that enables flexible, rechargeable solid-state batteries to be printed at industrial scale. These ultra-thin, printed batteries claim to be far more flexible, safer and more environmentally friendly than traditional metal-based batteries. Typical applications include low-cost IoT sensor labels for packaging and stock management, wireless industrial sensor technologies, and self-powered signage solutions. The technology can also be used with energy harvesting components to create self-powered signage, packaging and other innovative devices.

Otava announced that 5G millimeter-wave solutions provider TMY Technology has chosen Otava’s OTFL tunable band-pass filters for inclusion in its communications satellite subsystems. The OTFL301 claims to be the industry’s only digitally tunable band-pass filter IC with an operating frequency of 24 to 40 GHz and is designed for use in applications ranging from 5G and defense software-defined radios to electronic warfare, radar, signals intelligence, and instrumentation. An evaluation board allows designers to use a vector network analyzer to measure the filter’s performance. It includes a pre-programmed microcontroller that controls the bits of the filter and provides DC power and digital IO.

---