A big area of embedded system design are modules and there were plenty to see at this year’s embedded world conference. Modules support developers by helping reduce hardware design complexity, while accelerating development and time to market. These embedded modules come in a variety of flavors, depending on the application’s requirements, and are used across industries, including automotive, consumer electronics, industrial, and robotics as well as a range of IoT applications.
Many of the modules launched at embedded world fall into the categories of computer-on-modules (COMs), system-in-packages (SIPs), system-on-modules (SOMs), and single-board computers (SBCs). There also are a range of new wireless IoT modules, aimed at simplifying wireless development, particularly for companies that may lack RF expertise.
Here is a sampling of modules on display ate embedded world, designed to simplify hardware design in a range of applications.
Computer-on-modules
Adlink Technology Inc. showcased a variety of modules at embedded world. Two of the newest ones are the COM Express (COM.0 R3.1) and COM-HPC modules, based on Intel 13th Gen Core processors (code-named Raptor Lake) for edge and IoT multitasking, featuring high performance per watt and industrial-grade stability.
Offering high flexibility and scalability, the Express-RLP COM Express (COM.0 R3.1) Type 6 module offers up to 14 cores, 20 threads, 64-GB DDR5 SO-DIMM, PCIe Gen4 and an extremely rugged option at 15/28/45 W TDP. The COM-HPC-cRLS Client Type Size C module offers up to 24 cores, 128-GB DDR5 SO-DIMM, PCIe Gen5 and 2 × 2.5 GbE LAN, and it is also available in an extremely rugged option.
Adlink’s Express-RLP (Source: Adlink Technology Inc.)
The modules support Intel TCC and time sensitive networking (TSN), and are suited for hard-real-time computing workloads required by applications such as industrial automation, autonomous driving, AI robots and aviation, said Adlink.
Adlink also unveiled its MXM-Axe discrete GPU modules. The graphics card on the embedded MXM form factor integrates the latest Intel ARC A-Series graphics processor for Mobile, code-named Alchemist. It leverages Intel’s graphics ecosystems, such as OpenVINO for AI and Intel OneAPI (including Intel MediaSDK) management tools.
Other demos included AI at the edge for autonomous driving, machine vision and robotics, and Adlink’s latest Edge AI platforms and all-in-one AI camera development kit.
Congatec GmbH announced at embedded world that it is adding processors from Texas Instruments, Inc. to its Arm-based SMARC COMs. The first solution released is the conga-STDA4, a SMARC COM featuring the industrial-grade Arm Cortex-based TDA4VM processor. The system-on-chip (SoC) integrates accelerated vision and AI processing, real-time control and functional safety capabilities.
Congatec’s conga-STDA4 SMARC COM (Source: Congatec GmbH)
The dual Arm Cortex-A72-based module is designed for industrial mobile machinery, requiring near-field analytics, such as automated guided vehicles, autonomous mobile robots, construction equipment and agricultural machinery. Other applications include vision-focused industrial or medical solutions that require powerful but energy-efficient artificial intelligence (AI) processors at the edge.
Congatec said by integrating the TI TDA4VM processor on a standardized COM it simplifies the design-in process of the processor technology, which helps to reduce costs and shorten time to market.
First booting samples will be available in mid-2023. Series production is scheduled for 2024.
System-on-modules
Digi International has expanded its Digi ConnectCore MP1 family of SOMs with the launch of the Digi ConnectCore MP13. It is among the industry’s smallest SOMs with pre-certified Wi-Fi 5 and Bluetooth 5.2, housed in the Digi SMTplus 29 × 29-mm surface-mount form factor, according to the company.
Digi ConnectCore MP13 SOM (Source: Digi International)
The Digi ConnectCore MP1 SOMs are fully integrated solutions, which include a range of development tools, design support, wireless connectivity options, software and security features. It includes access to Digi ConnectCore Cloud Services and Digi ConnectCore Security Services.
The new wireless and secure SOM, based on the STMicroelectronics STM32MP133C MPU, targets applications for the medical, smart energy and industrial sectors, including industrial gateways, renewable energy controllers and electric vehicle (EV) charging stations. It incorporates the Digi TrustFence embedded security framework and fully validated Digi Embedded Yocto Linux software.
Other key features include pre-certified Wi-Fi 5 802.11a/b/g/n/ac + Bluetooth 5.2 (including DLE) option, up to 1-GB SLC NAND flash, up to 1-GB DDR3, dual 10/100/1000 Ethernet connectivity and ultra-low power and wake-up state management. It also comes with an off-the-shelf development board and low-cost gateway reference design.
At embedded world designers got a first look at Arduino’s new Portenta C33 IoT SOM, based on the RA6M5, an Arm Cortex M33-based microcontroller (MCU) from Renesas Electronics, targeting industrial IoT. The latest addition to Arduino’s Portenta family is specifically designed to provide a complete and scalable platform for IoT edge devices, packing a powerful MCU, connectivity and security at cost-effective prices. Applications include edge AI devices, IoT gateways, remote control systems, fleet management and process tracking.
Arduino’s Portenta C33 (Source: Arduino)
The Portenta C33 provides professional developers with a production-ready IoT platform, said Arduino, offering the Arduino environment with other high-level languages like MicroPython. The module offers Wi-Fi and Bluetooth LE connectivity and a variety of peripheral interfaces, including CAN, SAI, SPI and I2C. Touting industrial-grade security at the hardware level, the RA6M5 with Arm TrustZone integrates Secure Crypto Engine 9 (SCE9) with secure key storage. It also provides the capability to perform over-the-air (OTA) firmware updates with Arduino IoT Cloud or other third-party services.
The Portenta C33 form factor is compatible with Portenta and MKR ranges. It also can be paired with Nicla modules’ smart sensors for quality assurance controls on the production line to catch defects.
Arduino cited several applications for the Portenta C33, including industrial IoT gateways, connecting devices, machines and sensors to collect data in real time and display them on a custom Arduino IoT Cloud dashboard or third-party cloud platforms with end-to-end secure data encryption; smart systems to detect anomalies, alert operators about unplanned downtimes, or provide the information for predictive maintenance, and in buildings it can be used to monitor gas, water, electricity and any other utility consumption through a single system.
Portenta C33 will be commercially available by the end of the second quarter of 2023 at Arduino’s online store and through its authorized distributors. Visit https://www.arduino.cc/pro to join the Portenta C33 waiting list or find more information and complete technical specs.
Also unveiled at embedded world is Thundercomm America Corp.’s TurboX C2210, C4210 and C5430 SOMS for robotics and handheld devices. The new smart IoT modules target a variety of AIoT industry segments, also including tablets and smart cameras.
The TurboX C2210 SOM, powered by the Qualcomm QRB2210 SoC, provides discrete memory for more memory options and a system fast boot. It supports 1080P at 30 frames per second (fps) video decode and a long life cycle to 2028 and delivers advanced features to entry-level devices with better performance and improved power consumption. The TurboX C2210 SOM supports the Linux system and can be used in applications such as entry-tier robotics, e-scooter, smart home dock, smart camera and rugged handhelds.
The TurboX C4210 SOM, powered by the Qualcomm QRB4210 SoC, features triple ISPs for advanced dual-camera experience, dedicated compute DSP for computer vision and video post-processing. TI supports advanced graphics and high value features to mid-tier devices. Target applications include both industrial and consumer applications that require high data rates multimedia function, such as robotics, dash camera, video conference and smart retail.
The TurboX C5430 SOM integrates the Qualcomm QCS5430 SoC, a 6-nm processor with high performance, power efficiency, premium connectivity and 3.5 TOPS computing power. Packed with features, it supports 4K at 30 fps video decode and 4K at 30 fps video encode, Wi-Fi 6E, long range Bluetooth and four MIPI CSIs as well as a range of peripheral interfaces. It is a high performance AIoT SOM, targeting handheld devices, industrial robots and service robots.
The Qualcomm Robotics RB1 and RB2 development kits and SOMs for the TurboX C2210, C4210 and C5430 are available now for evaluation and testing.
Variscite demonstrated its newest SOMs for the first time at embedded world. These include i.MX93 and AM62x SOMs, aimed at AI/ML, industrial and cost-sensitive embedded devices.
The VAR-SOM-MX93 targets energy-efficient ML edge devices used in industrial, IoT and smart device applications. The SOM is powered by a 1.7-GHz dual Cortex-A55 NXP i.MX 93 processor, the industry’s first implementation of the Arm neural processing unit (NPU), Ethos-U65 microNPU; accelerates ML workloads and offers an energy-flex architecture for efficient processing, said the company. Other features include a 250-MHz Cortex-M33 real-time co-processor, AI/ML capabilities, built-in security and a range of industrial features. Prices start at $39.
The VAR-SOM-AM62, powered by Texas Instruments’ AM62x, targets cost-sensitive embedded products that require low power, high performance and a GPU. The SOM runs on the 1.4-GHz quad-core Cortex-A53 AM625x with a 400-MHz Cortex-M4F and additional 333-MHz PRU real-time co-processors. It offers a range of connectivity options, including a camera interface, dual LVDS display, certified dual-band Wi-Fi, BT/BLE 5.2, 3× CAN bus, dual USB and dual GbE. Prices start at $39.
Variscite released ML performance benchmarks for these new SOMs, comparing three modules with different CPU performance, pricing, power efficiency level and ML options. In internal benchmarking, running video-based object classification tests, the VAR-SOM-MX93 using the Ethos-U65 NPU performed with an average inference time of 4.9 ms and an average video throughput of 30 fps on a 640 × 480 input video. Running the same tests on the VAR-SOM-MX8M-NANO with no NPU resulted in an average inference time of 300 ms and average video throughput of 1 fps. The VAR-SOM-MX8M-PLUS using its NPU delivered an average inference time of 2.9 ms and an average video throughput of 30 fps.
Both new SOMs are compatible with the VAR-SOM Pin2Pin product family for easy scalability. It allows customers to use the same carrier board for all platforms and enables them to scale from entry-level to high-performance modules.
System-in-packages
Also aimed at robotics and IoT projects with Qualcomm SoCs, Lantronix Inc. introduced a new Open-Q family of SIP modules at embedded world. These include the entry-level Open-Q 2210RB and mid-tier Open-Q 4210RB SIP chipset modules as well as the companion Open-Q RB1 and RB2 development kits. The two SIP modules are footprint compatible and are offered in a Yocto Project environment for the development of embedded and IoT devices.
The SIP modules are powered by the Qualcomm QRB2210 and Qualcomm QRB4210 SoCs from Qualcomm Technologies Inc., which feature heterogeneous compute capabilities with multiple specialized processing cores for high-performance computing with lower power consumption.
The Open-Q 2210RB SIP compact (35 × 35mm) module is suited for the development of industrial IoT applications with improved power performance and increased graphic capabilities. The Open-Q 4210RB is a compact (36 × 36 mm) production-ready SIP chipset module that can be used in applications requiring AI intelligence, vivid graphics and dynamic camera capabilities.
The Open-Q RB1 and RB2 development kits support the SIP modules for rapid prototyping. Lantronix also offers engineering services to help with IoT product design.
Octavo Systems introduced its OSD62x family of SiPs, based on TI’s AM623 and AM625 processors, for edge and small form factor embedded processing. The new SIPs family targets small-size human-machine interface (HMI) applications for building automation, industrial control, IoT gateways and AI at the edge. It also is suited for other low-power systems that require high performance Linux processing.
Touted as the smallest AM62x module form factor, the OSD62x SiP family integrates functions such as high-speed memory, power management and passive components in a single BGA package. It reduces solution sizes by 60% and enables designs to get to market up to nine months faster, according to Octavo.
The OSD62x SiP family is offered in two package options. The first Maximized Integration (-MI) option, provides everything needed for an AM62x-based system in a 22 × 22 mm, 1-mm ball pitch BGA. The MI option includes the TI AM62x processor (up to 1.4 GHz Quad Arm Cortex-A53 with 400 MHz Cortex-M4F), LPDDR4 memory, power management, EEPROM, oscillator, passive components and a range of optional devices, including programmable logic, real-time clock (RTC) and analog-to-digital converter (ADC).
The second Size Optimized (-SO) package provides the smallest AM62x-based module form factor in a 14 × 9 mm, 0.5-mm ball pitch BGA. The SO option integrates the TI AM62x processor (up to 1.4 GHz Quad Cortex-A53 with 400 MHz Cortex-M4F), LPDDR4 memory and passive components.
The OSD62x SIPs are under development with samples expected by the end of 2023. Contact Octavo if you are interested in joining the OSD62x SIP beta program.
Recommended
Embedded world 2023: MPUs, MCUs, SoCs and RISC-V
Single-board computers
Targeting industrial IoT, Asus IoT announced at embedded world its new Tinker V SBC powered by a 64-bit RISC-V-based processor, which supports both Linux Debian and Yocto operating systems. The RISC-V processor uses the open-source Instruction Set Architecture (ISA), based on Reduced Instruction Set (RISC) principles. It is the company’s first open-source RISC-V SBC for industrial IoT developers.
Asus IoT Tinker V SBC (Source: Asus IoT)
Packing a lot of functionality and connectivity into a compact Pico-ITX form factor, the Tinker V features the Renesas RZ/Five MPU, which incorporates the RISC-V AndesCore AX45MP single-core, supporting 1.0-GHz operating frequencies. It supplies a range of peripheral connectors for industrial use, including GPIO, micro-USB, dual gigabit Ethernet, a pair of CAN bus interfaces and two RS232 COM ports. Other features include 1 GB of built-in RAM, an optional 16 GB eMMC and a wide operating temperature range of -20°C to 60°C.
Asus IoT Tinker V samples will be available in Q2. Customers will receive at least five years of support and dedicated on-site technical support.
Also targeting industrial applications as well as robotics, edge computing and AI vision systems, DFI launched its PCSF51 next-generation 1.8″ Pi SBC at embedded world. The SBC is powered by the AMD Ryzen Embedded R2000 processor, offers low power consumption and features advanced graphics processing and computing capabilities in a small form factor. Thanks to its scalability with various expansion ports for a range of I/O interfaces including HDMI, USB 3.1 and Gigabit Ethernet connectivity, the SBC can be used in a variety of applications.
DFI’s PCSF51 SBC (Source: DFI)
The PCSF51 offers CPU options for 4-GB or 8-GB single-channel DDR4 memory and eMMC storage capacities of up to 128 GB. Compared to its previous generation GHF51 with the AMD Ryzen R1000 Processor, the latest R2000 processor offers upgraded performance with double the max CPU core count, scalability up to four CPU cores and eight threads and a 50% increase in CPU power, according to DFI. In addition, the graphics performance is increased by 15% with up to eight graphic compute units.
DFI also made it easier to use the on-board power/reset and HDMI, while thinning down the thermal module. It also supports ultra-high-resolution displays in 4K resolution.
Wireless modules
U-blox has launched its first tri-radio standalone module with dual-band Wi-Fi 6, Bluetooth LE and Thread (including support for Matter). With the three radios, the module can be used in a variety of use cases and provide integrated coexistence between radios, delivering less radio interference, a lower bill of materials and easier integration.
U-blox’s IRIS-W1 tri-radio module (Source: u-blox)
The u-blox IRIS-W1 is available in four variants, delivering a selection of antennas and software environments. Thanks to the embedded Arm Cortex-M33 MCU and high level of security, the module can be implemented in a variety of applications. These include industrial automation, medical equipment, EV charging, smart buildings, smart homes (including intelligent appliances), after-market telematics and industrial power tools.
In particular, smart homes and smart buildings will be able to leverage the modules support for Matter over Thread, Wi-Fi and Ethernet for faster development.
U-blox said it is breaking new ground with its first dual-band Wi-Fi 6 standalone module, built on the NXP RW612 chipset, offering a range of advantages for IoT devices. Wi-Fi 6 reduces power consumption and increases network efficiency and the number of served clients; enables users to select Wi-Fi channels from the 2.4 and 5 GHz spectrum, and it is compatible with battery-operated devices.
In addition, the embedded Arm Cortex-M33 MCU eliminates the need for an external processing unit and with the available RAM and Flash memory, designers can leverage the processing capabilities for advanced applications, said u-blox. It also includes advanced security featuring the NXP EdgeLock and Arm TrustZone.
Other features include a range of peripheral interfaces, including UART, USB, SPI, SDIO, RMII, QVGA, I2S, I2C and GPIOs, and a miniature form factor. The IRIS-W101 measures 14.6 × 16.8 × 2.1 mm, while the bit larger IRIS-W106 comes in a 14.6 × 20.9 × 2.1-mm footprint.
Also targeting a range of IoT applications is a Wi-Fi HaLow module from Quectel Wireless Solutions. Launched at embedded world, the FGH100M Wi-Fi HaLow module provides long range data transmission, low power consumption and less design complexity. The company said the module delivers improved penetration through walls and other obstructions for a wide range of indoor and outdoor IoT applications.
Quectel’s FGH100M Wi-Fi HaLow module (Source: Quectel Wireless Solutions)
The FGH100M module, supporting the IEEE 802.11ah wireless networking protocol, also known as Wi-Fi HaLow, meets IoT requirements with its sub-1 GHz signal coverage, which allows users to control IoT devices in a one-kilometer radius, said the company. It also provides ten times the range of traditional Wi-Fi, making it suited for both indoor and outdoor IoT environments such as home and industrial automation, smart agriculture, smart city, smart building, warehouse, retail store, and campus.
The wireless module, based on Morse Micro’s MM6108 Wi-Fi HaLow SoC, is compliant with IEEE 802.11ah Wi-Fi standard and operates in the 850 – 950 MHz bands with 1/2/4/8 MHz channel width. It features 21-dBm maximum output power and 32.5 Mbits/s maximum transmission rate.
Quectel said the Wi-Fi HaLow module enables lower power connectivity than previous Wi-Fi technologies and supports coin cell battery devices operating for months or years, and unlike similar IoT technologies, Wi-Fi HaLow supports native IP and eliminates the need for proprietary gateways, controllers or hubs. This translates into streamlined installation and a reduction in operating costs.
The FGH100M meets Wi-Fi requirements for authentication and encryption including WPA3 and AES for OTA traffic. It also supports SHA-256, SHA-384 and SHA-512 encryption algorithms to better protect transmission data and offers a range of interfaces including UART, SPI, I2C, SDIO 2., and PWM.
The FGH100M Wi-Fi HaLow module, housed in a 13.0 × 13.0 × 2.2-mm footprint, is available now in engineering samples.
Related embedded world product launches
Infineon MCUs support Rust programming language
Quick-Connect Studio speeds IoT design
Wi-Fi + Bluetooth combo chip extends battery life
Smart BMS supports 20-cell battery packs
Connected MCU enables secure NFC authentication
Microchip adds six secure authentication ICs
Automotive MCU targets wireless applications
Advertisement
