Telit Cinterion and Alif Semiconductor have launched the Vision AppKit, a smart connected camera reference design that delivers high efficiency processing in a small form factor for object detection and image classification. Described as a postage stamp-sized, intelligent connected camera platform, the reference design combines Telit Cinterion’s Wi-Fi and Bluetooth wireless technology or LTE-M communication modules with Alif’s Ensemble E3 microcontroller (MCU) series, which can perform on-device AI use-cases like face and object detection and image classification at lower power consumption than previously possible, according to the companies.

The reference design, powered by the E3 Series MCU with edgeAI capabilities for battery-operated IoT devices, is an ultra-compact and ultra-low-power AI-enabled camera that can capture images and/or video, perform AI-based data processing in real-time and deliver the results wirelessly to a display or other external system.

“This technology enables battery-operated cameras to perform computations previously only possible in the cloud,” said Reza Kazerounian, co-founder and president at Alif Semiconductor, in a statement.

Atmosic Technologies has introduced its ATM34/e high-performance system-on-chip (SoC) series with multiprotocol support. Building on its ATM3 and ATM33/e SoCs with Bluetooth LE support, the ATM34/e series adds IEEE 802.15.4-based protocols including Thread and Matter, as well as enhanced Bluetooth LE for the latest 5.4 standard revision.

“Interoperability in the smart home is critical, creating a growing demand for connected devices that support standards-based wireless protocols,” said David Su, Atmosic’s CEO, in a statement.  “Thread and Matter enable smart home and mobile platforms to seamlessly communicate with each other regardless of the device manufacturer.”

Targeting ultra-low-power end IoT devices such as sensors and switches, the ATM34/e series typically uses 50% less active energy than competitor solutions, according to Atmosic. The SoCs incorporate the company’s radio that consumes less than 1 mA in receive mode and an energy-efficient transmitter with an output power range from -20 dBm to +10 dBm. These power-saving features extend battery life in connected devices and can enable some devices to operate battery-free.

Building energy-harvesting capabilities into the wireless SoC, the ATM34/e series enables devices to capture, use and store energy from RF, heat, light and motion sources. It offers a wide energy harvesting input range and supports a range of storage devices, which enables smart sensors and other low power automation applications to operate unattended for years, the company said.

The ATM34/e series also offers a new scalable memory architecture, which allows designers to select a memory footprint that meets their protocol, application and cost requirements. Key features include a 64-MHz Arm Cortex M33F processor to optimize power and performance and support for Arm’s TrustZone technology for security.

The ATM34/e series is currently available in two packages: a 5×5 40-pin QFN and a 7×7 56-pin QFN. Both packages are pin-to-pin compatible with Atmosic’s ATM33/e series Bluetooth SoCs for easy upgrading. Product samples will be available starting in the second quarter of 2024. Click here for more product information. Atmosic is showcasing its products at CES 2024 at the Venetian Palazzo.

Micron Technology, Inc. has announced the industry’s first standard low-power compression attached memory module (LPCAMM2), targeting the replacement of SODIMMs in PCs with higher performance, better energy efficiency, space savings and modularity. Announced at CES 2024, the LPDDR5X-based LPCAMM2, available in capacities from 16 GB to 64 GB, is sampling now, followed by production in the first half of 2024.

Micron said LPCAMM2 is the first disruptive new form factor for client PCs since the introduction of SODIMMs in 1997. With the LPDDR5X DRAM incorporated into the LPCAMM2 form factor, the memory device is reported to provide up to 61% lower power, up to 71% better performance for PCMark 10 essential workloads such as web browsing and video conferencing and  64% space savings over SODIMMs.

LPCAMM2 is expected to meet the performance requirements of generative AI applications to process AI workloads on PCs and scale to applications that need a higher performance and lower power form factor as well as the ability to upgrade low-power DRAM, according to Micron.

“This first-of-its-kind product will enhance the capabilities of AI-enabled laptops, whose memory capacity can be upgraded as technology and customer needs evolve,” said Praveen Vaidyanathan, vice president and general manager of Micron’s Compute Products Group, in a statement.

Micron in collaboration with PC OEMs and ecosystem partners helped design and develop the LPCAMM2 form factor. The new memory type also has resulted in several innovations for test hardware, testing methodologies and automation technologies, Micron said.

Key features include:

    Higher performance with LPDDR5X to achieve speeds up to 9600 Mbits/s versus 5600 Mbits/s with current DDR5 SODIMMs
    Up to 80% system standby power savings to improve battery life
    Up to 7% better performance for digital content creation workloads
    Up to 15% improvement for productivity workloads in PCMark 10 tests

The LPCAMM2 form factor also features a single PCB for all module capacities, delivering supply chain benefits to OEMs and ODMs, Micron said. It also improves routing complexity with a simplified motherboard compared to SODIMM.

Micron will also offer Crucial LPCAMM2 memory solutions that allow users to upgrade their memory themselves, cited as an industry first for low-power memory thanks to the form factor’s upgradeability. Crucial LPCAMM2 products will be available in the first half of 2024.

NXP Semiconductors N.V. has expanded its automotive radar one-chip family, introduced this time last year, with the launch of the new SAF86xx. The device monolithically integrates a high-performance radar transceiver, a multi-core radar processor and a MACsec hardware engine for secure data communication over automotive Ethernet.

The radar one-chip, together with NXP’s S32 high-performance processors, vehicle network connectivity and power management, delivers a full system solution that readies the path for software-defined radar, from smart sensors to future streaming sensors, NXP said. This includes 360-degree sensor fusion, better sensor resolution and AI-based object classification.

The highly integrated 76 to 81 GHz radar SoC is used for streaming low-level radar sensor data at up to 1 Gbit/s. NXP said it will help carmakers optimize next-generation ADAS partitioning for software-defined vehicles and provide a smooth transition to new architectures. It also will allow OEMs  to introduce new software-defined radar features during the lifetime of the vehicle with over-the-air (OTA) updates.

“Using our new SAF86xx radar one-chip family, OEMs can quickly and easily migrate their current radar platforms to new software-defined vehicle architectures,” said Steffen Spannagel, SVP and GM, ADAS, NXP Semiconductors, in a statement. “A network of connected radar sensors with software-defined functions on a dedicated S32R radar processor in a distributed architecture can enhance radar-based perception to support advancements in autonomous driving. That includes 360-degree sensing, more powerful AI-based algorithms and secure OTA software updates.”

The SAF86xx  shares a common architecture with the SAF85xx introduced last year and also leverages 28-nm RFCMOS performance for improved radar sensor capabilities, compared to prior-generation 40-nm or 45-nm products. Other features include an extended detection range beyond 300 meters and more reliable detection of small objects like curb stones as well as vulnerable road users including cyclists and pedestrians.

The SAF86xx supports NCAP safety functions and advanced ADAS and autonomous driving applications, including advanced comfort features for SAE levels 2+ and 3. It is developed in accordance to ISO 26262 Safety Element out of Context (SEooC) methodology supporting ASIL Level B, in accordance to ISO/SAE 21434 (as a component-out-of-context) and meets the latest security requirements through its HSE security engine

The SAF8xxx family, featuring the SAF86xx and SAF85xx, can be tailored for OEM applications. It supports a range of sensor outputs, including object, point cloud-, or range-FFT-level data for smart sensors in today’s architectures and streaming sensors in future distributed architectures, NXP said. It is sampling now for alpha customers.

NXP also announced that automotive electronics supplier HELLA will use NXP’s SoC family for its 7th generation radar portfolio, including front, rear, corner and side radar. NXP’s radar portfolio is on display at CES 2024, LVCC, booth CP-18.

Rambus Inc. has announced the availability of the industry’s first Gen4 DDR5 registering clock driver (RCD). The Gen4 DDR5 RCD enables a 50% increase in memory bandwidth and increases the data rate to 7200 MT/s from the first-generation 4800 MT/s DDR5 module solutions, enabling a new level of main memory performance for data-center servers, according to the company.

The improvements meet the requirements for advanced data-center server and generative AI workloads. “With memory being an essential enabler of server performance, the need for greater memory bandwidth continues its meteoric rise driven by demanding workloads like generative AI,” said Sean Fan, chief operating officer at Rambus, in a statement.

The DDR5 RCD provides command/address (CA) and clocks to the DDR5 memory devices in RDIMMs. The device supports the double data rate (DDR) and single data rate (SDR) CA bus and two independent subchannels per RDIMM. It provides clock rates up to 3600 MHz and supports data rates up to 7200 MT/S. Other features include low power 1.1-V VDD operation, four clocks per subchannel and a 12.5 MHz (max) I3C bus interface.

The company’s DDR5 memory interface chips provide key components of server systems, including the RCD, serial presence detect (SPD) hub and temperature sensors. These devices target advanced data-center applications such as hyperscale data centers, high-performance computing (HPC), high-capacity storage systems, big data and real-time analytics.

The Rambus 7200 MT/s DDR5 RCD is available now. Sampling to the major DDR5 memory module (RDIMM) manufacturers started in the fourth quarter of 2023.

Adlink Technology Inc. has launched its cExpress-MTL, a compact COM Express Type 6 module powered by the new Intel Core Ultra processor, integrating the CPU, GPU and NPU for optimized performance and efficiency. The new COM Express module provides up to eight GPU Xe-cores (128 EUs), an NPU and 14 CPU cores at 28 W TDP, delivering 1.9× the GPU performance than the previous generation and dedicated AI acceleration with lower power consumption.

“As demands for varying workloads continue to grow in battery-powered edge applications, which cannot solely rely on the CPU alone, we are excited to showcase our Intel Core Ultra-powered COM Express,” said Dylan Cheng, senior product manager at Adlink COM, in a statement. “By integrating a GPU with 1.9× higher performance than prior generation and an NPU for dedicated AI acceleration on top of its powerful CPU, the featured processor in our cExpress-MTL offers proficient graphics and AI capabilities without needing to connect to additional processing units. This substantially simplifies your solution designs, translating to accelerated time-to-market and lower total cost of ownership.”

With the Lower Power E core to handle light workloads, the Intel Core Ultra processor can be 30% to 50% more power efficient than 13th Gen Intel Core, enabling high performance even in compact, battery-powered designs, according to Adlink. Applications include portable medical ultrasound, industrial automation, autonomous driving and AI robots.

Thanks to the hardware-accelerated AV1 encoding and decoding functionality, the cExpress-MTL module can deliver instantaneous media streaming with minimal latency. Other key features include PCIe interfaces upgraded to Gen 4 for faster data transmission, USB4 support and 2.5 Gigabit Ethernet with optional TSN.

Adlink said it is working on I-Pi development kits based on the cExpress-MTL module for out-of-the-box-ready prototyping and referencing.

Building on its recent Snapdragon XR2 Gen 2 platform, Qualcomm Technologies, Inc. has unveiled the Snapdragon XR2+ Gen 2 chip architecture that delivers 4.3K spatial computing at 90 frames per second (fps) for more realistic mixed reality (MR) and virtual reality (VR) experiences. The single-chip architecture is designed to deliver crisper and more detailed visuals in thinner and more comfortable headsets.

In addition to the 4K resolution (4.3K × 4.3K per eye), the “+” platform delivers 15% higher GPU and 20% higher CPU frequency, compared to the previous generation, with faster frequency rates and less jitter for crisper visuals. It also supports 12 or more concurrent cameras and on-device AI to track the user’s head, hands, eyes and controllers for higher precision navigation. It also includes an upgraded Spectra ISP and 12-ms full color video see-through.

Like the Snapdragon XR2 Gen 2 platform, the + platform offers 2.5× higher GPU performance, up to 50% greater GPU power efficiency and an 8× improvement in AI performance. It also supports Wi-Fi 7, Wi-Fi 6E and Bluetooth 5.3 and 5.2 wireless connectivity. The Snapdragon XR2+ will power Samsung’s Galaxy and Google XR devices.

Qualcomm Technologies also offers a new MR and VR reference design developed by Goertek, which features Tobii’s eye tracking technology. This reference design supports both 3K per eye (XR2 Gen 2) and 4K per eye (XR2+ Gen 2) configurations.

Asahi Kasei Microdevices (AKM) unveiled its millimeter-wave (mmWave) radar transceiver technology with high-resolution detection at CES 2024. Developed in partnership with Pontosense, the AK581x 60-GHz mmWave radar transceiver addresses two detection challenges in automotive and home safety applications.

The radar transceiver IC delivers accurate, high-resolution detection with automotive-level reliability for wirelessly detecting children left behind in cars within seconds, aiming to surpass current regulations and standards for child presence detection (CPD) worldwide. In addition to child detection, older adults regularly have serious complications from falls, and the radar-based solution supports rapid emergency response. The technology also has applications in home automation, energy efficiency and public safety in smart city infrastructures.
AKM’s mmWave radar transceiver module installed in a demo car, compared with an industry-standard module (green).

The AK581x radar transceiver features a multi-channel architecture with four receivers and four transmitters and supports a frequency modulation range to 7 GHz (57–64 GHz operating frequency range). This allows for 50-cm isolation performance and a high range resolution as fine as 2.2 cm. Thanks to these features, it provides the capability to detect multiple people separately within a 1-meter range and the simultaneous detection of the respiratory rates and positions of multiple people.

When the radar technology is paired with Pontosense’s artificial intelligence (AI) and machine learning (ML) software, the AK581x series delivers the flexibility of different levels of resolution and fields of view, without the need for cameras and microphones.

The AK581x series is available in automotive grade (AK5818) and consumer grade (AK5816) versions. The AK5816 targets consumer electronics and IoT applications. Key features include an integrated selector, enabling the 4Tx/4Rx time division operation to be handled by a single IC for lower power consumption; a built-in radar signal processing function that allows for semi-automatic continuation of radar operation by pin setting, and support for signal processing output via serial communication.

The automotive AK5818 also features four transmit and four receive channels with time-division operation. Digital signal processing is also available. Key features include an all-digital PLL in synthesizer, wideband delta-sigma AD converter, MIPI CSI-2 output, transmit power adjustment function, baseband signal amplitude adjustment and serial interface (SPI or I2C).

Both devices are housed in 115-pin BGA packages. Mass production of the AK5816 is scheduled for the second quarter of 2024, followed by the AK5818 in the fourth quarter of 2024. AKM will demo the radar ICs at CES 2024, LVCC North Hall, Booth #11051.

Murata Manufacturing Co., Ltd. has introduced its new Type 2FR connectivity module, driven by NXP Semiconductor’s RW612 wireless microcontroller (MCU) with an integrated tri-radio. Measuring 12 × 11 mm, the tri-radio module is 50% smaller than a discrete implementation. It delivers high integration, efficiency and tri-radio capabilities that Murata claims is a breakthrough in IoT connectivity. Applications include smart home devices, smart appliances, enterprise and industrial automation as well as smart city and smart energy solutions.

The module supports communication protocols such as dual-band Wi-Fi 6, Bluetooth Low Energy (LE) 5.3, 802.15.4 and Ethernet. Importantly today, Type 2FR is ready for Matter, allowing compatibility with Matter over Wi-Fi, Matter over Thread and Matter over Ethernet.

The Wi-Fi and Ethernet capability allows the IoT device to function as a border router to address new applications, according to the company.

The module features a 260-MHz Arm Cortex-M33 core, 1.2-MB RAM, 16-MB flash memory and NXP EdgeLock security to meet a broad spectrum of IoT applications. The module  can operate in co-processor mode. The RW612 is supported by the NXP MCUXpresso ecosystem of software and tools. The solution has full regulatory certification and offers multiple external antenna options for compliance and reliability.

Murata will showcase its Matter-enabled tri-radio module portfolio at CES 2024, LVCC, West Hall, Booth #6300. Samples will be available at the same time.

Texas Instruments Inc. (TI) has debuted three new automotive chips at CES 2024 that are designed to improve automotive safety. One of those chips is the AWR2544 77-GHz millimeter-wave (mmWave) radar sensor, claimed as the industry’s first for satellite radar architectures. This satellite radar chip enables higher levels of autonomy by improving sensor fusion and delivering more accurate decision-making in advanced driver assistance systems (ADAS).

Rounding out the product launch are two software-programmable driver chips, the DRV3946-Q1 integrated contactor driver and the DRV3901-Q1 integrated squib driver for pyro fuses, targeting advanced battery management systems (BMS). They offer built-in diagnostics and support functional safety for battery management and powertrain systems. TI is demonstrating the new automotive chips at CES.

The AWR2544 single-chip radar sensor can increase vehicle sensing ranges beyond 200 meters and improve ADAS decision-making. “In satellite architectures, radar sensors output semi-processed data to a central processor for ADAS decision-making using sensor fusion algorithms, taking advantage of the 360-degree sensor coverage to achieve higher levels of vehicle safety,” TI said.

Satellite radar architectures enhance ADAS decision-making by centrally processing radar data from multiple sensors, which improves radar performance, TI said. It also enables radar scalability for ADAS systems with different performance levels and number of sensors around the car.

The AWR2544 automotive satellite radar chip is also the industry’s first with launch-on-package (LOP) technology, which can reduce the size of the sensor by as much as 30% by mounting a 3D waveguide antenna on the opposite side of the printed circuit board, TI said. LOP technology extends sensor ranges beyond 200 meters with a single chip, and in satellite architectures, it enables ADAS systems to make “smarter decisions from farther away.” (Read Are you ready for the emerging automotive radar satellite architecture? for more information.)

The new driver chips support software-defined vehicles by contributing to smarter and more efficient battery management. The two new highly integrated driver chips target requirements for safer and more efficient control of high-voltage disconnect circuits in a BMS or other powertrain system. Both drivers are functional safety-compliant to ISO 26262.

For BMS and other powertrain systems, TI claims the DRV3946-Q1 is the industry’s first fully integrated contactor driver. Key features include a peak-and-hold current controller and low-ohmic power stage that help increase system power efficiency and safety diagnostics to monitor the condition of the contactor. It also supports a range of solenoid types with flexible current control parameters.

The DRV3901-Q1 squib driver delivers an intelligent pyro fuse disconnect system thanks to built-in circuitry that monitors the pyro fuse and provides diagnostic information to the system microcontroller. The pyro fuses reduce design complexity in hybrid and electric vehicle (HEV/EV) BMS systems by replacing discrete circuitry with a single driver. It also improves intelligence and reliability of the disconnect signals by using inputs such as temperature and current signals, TI said. For more information, read How squib and contactor drivers help improve safety and efficiency in HEV/EV battery disconnect systems.

Preproduction quantities of the AWR2544, DRV3901-Q1 and DRV3946-Q1 and evaluation modules (DRV3901-Q1 and DRV3946-Q1) are available for purchase on TI.com.

TI is showcasing its automotive products at CES with demonstrations for ADAS and BMS designs as well as other technologies for applications such as zone architectures, EV charging and energy storage. Visit TI.com/CES for more information.