Virtium launches industry’s first SOM to include on-board hardware AI accelerator chip

Virtium Embedded Artists has announced the launch of the iMX8M Mini DX-M1, a system-on-module (SOM) which integrates a quad-core application processor and a 25 TOPS AI hardware accelerator chip with associated memory on a board with a compact 82mm x 50mm footprint.

By integrating a DEEPX DX-M1 AI accelerator into the SOM, Virtium Embedded Artists eliminates the need for embedded device manufacturers that want to implement AI functions to plug a discrete AI processor module into their system. This provides multiple benefits to OEMs and designers, including board space savings, simplified systems design, streamlined bill-of-materials, and faster time-to-market.

The new Virtium Embedded Artists SOM is based on the i.MX 8M Mini from NXP Semiconductors, an application processor which features four 1.6GHz/1.8GHz Arm® Cortex®-A53 CPU cores backed by 2GB of LPDDR4 memory, as well as a 400MHz Cortex-M4 controller core. The i.MX 8M Mini provides strong support for video and imaging applications thanks to its video engine with 1080p codec, 2D/3D graphics engine, and 4-lane MIPI-DSI interface, as well as a Gigabit Ethernet interface.

Working in tandem with the DEEPX DX-M1 NPU, the i.MX 8M provides the image processing capability required to support AI-enabled vision systems used in popular and emerging vision applications. These include:
• Drones
• Security and surveillance
• Automated inspection and monitoring
• Transportation
• MedTech
• AgTech

The DX-M1, a high-performance AI accelerator, provides 25 TOPS of throughput at an average power consumption of 5W, making the SOM ideal for power-constrained edge computing applications. In this new Virtium Embedded Artists implementation, the SOM provides 4GB of LPDDR5 memory for the AI processor accessed via a 64-bit, 4-channel data bus. This enables the DX-M1 to run multiple AI models concurrently without performance degradation.

Virtium Embedded Artists supplies the iMX8M Mini DX-M1 SOM in two versions: one with the DEEPX AI accelerator and i.MX 8M Mini processor, and one with the i.MX 8M Mini alone. This allows OEMs to easily scale up embedded computer systems based on the i.MX 8M Mini to also provide vision AI capabilities without altering their carrier board design.
Each of the two versions of the SOM is available with a 1.8GHz i.MX 8M Mini, for operation at temperatures between 0°C and 70°C, or with a 1.6GHz i.MX 8M Mini with an operating-temperature range of -40°C to 85°C.

Alongside the iMX8M Mini DX-M1 hardware, Virtium Embedded Artists provides a full technical support service. It also supplies the DEEPX iM8 Mini AI Kit, which includes reference design implementations for key interfaces. Virtium Embedded Artists provides pre-designed standard carrier boards to ease integration into customers’ designs.

https://www.embeddedartists.com

> Read More

Microchip enhances their digital signal controller lineup

Evolving security and functional safety demands, coupled with the growing complexity of real-time embedded applications, are driving designers to seek innovative solutions that deliver greater accuracy, improved reliability and compliance with industry standards. To address these challenges, Microchip Technology has added the dsPIC33AK512MPS512 and dsPIC33AK512MC510 Digital Signal Controller (DSC) families to its dsPIC33A DSC product line. The devices enable the implementation of computation-intensive control algorithms for improved energy efficiency in motor control, AI server power supplies, energy storage systems and complex sensor signal processing with Machine Learning (ML)-based inferencing.

The dsPIC33AK512MPS family delivers precise, high-speed control through 78 ps high-resolution Pulse Width Modulations (PWMs) and low-latency 40 Msps ADCs, enabling fast and accurate control loops essential for optimising the performance of Silicon Carbide (SiC) and Gallium Nitride (GaN)-based DC-DC converters. Additionally, dsPIC33AK512MPS devices include advanced security features, an integrated touch controller and a high pin count of up to 128 pins. The dsPIC33AK512MC family is designed to offer low-latency, 40 Msps ADCs and 1.25 ns PWM resolution, providing a feature- and cost-optimised solution for multi-motor control and complex embedded applications.

The dsPIC33A DSC families, with up to 512 KB Flash and a rich peripheral set, integrate a double precision floating-point unit to accelerate mathematical computations and leverage a 32-bit architecture for seamless adoption of model-based design code. Their enhanced instruction set and Digital Signal Processing (DSP) capabilities, including single-cycle MAC operations and a 200 MHz core speed, make these devices highly efficient for low-latency, real-time control applications. Supported by MPLAB Machine Learning Development Suite, dsPIC33A devices streamline the ML workflow by automating data preparation, feature extraction, training, validation and firmware conversion of optimized models.

With a range of hardware safety features, dsPIC33AK512MPS/MC DSCs are compliant with functional safety standards and are developed in accordance with International Organisation for Standardisation (ISO) 26262 and International Electrotechnical Commission (IEC) 61508 processes, making them suitable for safety-critical automotive and industrial applications. To further enhance system-level security, the dsPIC33AK512MPS DSC family includes integrated crypto accelerators and a Flash security module, enabling immutable root of trust, secure boot, secure firmware upgrades and secure debug capabilities.

https://www.microchip.com

> Read More

Tria’s new family of compute modules powered by Qualcomm processors

The new family of embedded compute modules from Tria, an Avnet company specialising in manufacturing embedded compute boards, not only have Qualcomm Dragonwing processors on board—they now support different operating systems. These options include Android, Windows 11 IoT Enterprise, and Yocto Linux. The variety of operating systems supported enable the modules to be used for embedded designs in the industrial, medical, agriculture and construction sectors, as well as any embedded application that can benefit from edge computing, machine learning and AI.

At the heart of these modules are high-performance, low-power Qualcomm Dragonwing processors, providing advanced technology in customised solutions. With advanced edge AI and seamless networking capabilities, Qualcomm Dragonwing targets embedded systems for various industries. These processors have been designed for speed, scalability and reliability, enabling customers to achieve smarter decision making, better efficiencies and to reach the market faster.

This means that Tria’s new lineup offers a powerful CPU and AI combination at lowest power consumption, enabling embedded designs to benefit from higher performance and increased capabilities.

Tria also now provides Microsoft Windows with ARM, allowing a transition from the x86 architecture to ARM with the same OS. This transition was previously not possible, since Windows on ARM was not available.

“Our collaboration with Qualcomm Technologies has enabled us to deliver high performance and low power technology with highest quality they are used to receiving from Tria in our new compute modules,” said Christian Bauer, Product Marketing Manager, Tria Technologies. “This family of modules is the only one available to offer compatibility with multiple operating systems at a competitive price.”

“With multi-OS support across Tria’s new modules powered by Qualcomm Dragonwing and Snapdragon platforms, customers gain the flexibility of Windows, Android and Linux across our different platforms”, said Douglas Benitez, Senior Director, Business Development, Qualcomm Europe, Inc. “Our collaboration with Tria is accelerating industrial innovation—delivering high performance, low power compute and accelerated AI at the edge.”

Qualcomm Dragonwing processors are known for their performance, efficiency and advanced AI capabilities, offering powerful performance in multi-core tasks and great GPU performance, but also more efficient power consumption in mobile and portable systems.

https://www.Tria-technologies.com

> Read More

ASRock launches industrial motherboards for AI-driven Edge applications

ASRock Industrial has announced its next generation industrial motherboards powered by Intel Core 200S series processors (Bartlett Lake-S), supporting up to 24 cores and 32 threads. The new series delivers improved single-thread, multi-thread and AI-performance, along with DDR5 5600 MHz memory support up to 192GB, advanced I/O capabilities, PCIe Gen 5/4 expandability, 4K quad-display powered by Intel UHD Graphics 770 with Xe architecture. With BIOS-level update support, the new series enables customers to seamlessly upgrade existing 14th, 13th, and 12th Gen Intel CPU-based motherboards to the latest generation. Available in Mini-ITX, Micro-ATX, and ATX form factors with longevity support, the motherboards are built for intelligent edge applications across smart manufacturing, autonomous robotics, machine vision, retail, kiosks, digital signage, gaming, and security environments.

ASRock industrial’s Intel Core 200S DDR5 supported motherboards are available in Mini-ITX, Micro-ATX, and ATX form factors, built on Intel W680, Q670, and H610 chipsets to meet a broad range of industrial computing demands.

For compact systems, the Mini-ITX models feature two DDR5 4800/5600 MHz memory slots up to 96GB, support quad displays on Q670/W680, triple displays on H610, and offer dual LAN ports—one 2.5G and one 1G. Models with streamlined I/O options include the IMB-1239-WV, IMB-1240-WV, IMB-1244-WV, and IMB-X1240-WV, IMB-X1244-WV, supporting H610, Q670, and W680 chipsets. For applications that require expanded I/O capabilities, additional Mini-ITX options such as the IMB-1237, IMB-1238, and IMB-X1238 are also available with the same chipset support.

In the Micro-ATX lineup, IMB-X1316-10G and IMB-X1316 are powered by the W680 chipset and IMB-1316 with a Q670 chipset support quad-display output, up to 192GB DDR5 4400/5600 MHz memory, multiple PCIe Gen5/Gen4 slots, and advanced networking with 10G (IMB-X1316-10G), 2.5G, and 1G LAN. IMB-1318 and IMB-1315, based on H610, featuring DDR5 4800/5600 MHz memory up to 96GB, triple HDMI (IMB-1318), dual HDMI (IMB-1315) and dual 1G LAN.

For high-performance needs, the ATX models includes IMB-X1714 with W680 chipset, IMB-1714 based on Q670 chipset, and IMB-1713 featuring H610 chipset. These ATX boards offer support up to 192GB DDR5 memory (IMB-1714/IMB-X1714), multiple PCIe Gen5/Gen4 slots, triple displays, and up to triple LAN with 2.5G support (IMB-1714/IMB-X1714). The series is ideal for smart manufacturing, autonomous robotics, AI inference, Retail and advances edge computing.

ASRock Industrial also provides a lineup of DDR4 supported motherboards for Intel Core 200S Series Processors, delivering solutions optimised for AI-enabled industrial environments.

Within the Mini-ITX high-rise I/O category, models such as IMB-1230, IMB-1241, and IMB-1242 are built on the H610 chipset, while IMB-1231 and IMB-X1231 adopt the Q670 and W680 chipsets respectively, supporting up to 64GB DDR4 3200 MHz memory, quad or triple displays outputs, and dual LAN for reliable edge connectivity. For compact thin I/O designs, the Mini-ITX series includes IMB-1232-WV with H610, IMB-1233-WV with Q670, and IMB-X1233-WV with W680, offering wide voltage input, essential I/O, dual LAN and up to quad display support, ideal for space-constrained deployments such as retail automation, kiosks, and edge devices running lightweight AI workloads.

The Micro-ATX models include IMB-1314 with Q670 and IMB-X1314 with W680, providing dual DDR4 3200 MHz memory support up to 128GB, four PCIe Gen4 slots, quad-display output, dual 2.5G LAN (IMB-1314) and triple 2.5G LAN (IMB-X1314), and USB 3.2 Gen2 for smooth data processing and connectivity in AI-assisted manufacturing and system monitoring applications.

In the ATX form factor, IMB-1712 powered by Q670 and IMB-X1712 by W680 offer high expandability with triple 2.5G LAN (IMB-X1712,) and dual 2.5G LAN (IMB-1712), four PCIe Gen4 slots, M.2 and SATA storage interfaces, and triple display output including HDMI, DP and VGA. These platforms are designed for long-life deployment in industrial control, intelligent surveillance, smart retail, and AI-enhanced display centric systems at the edge.

ASRock Industrial’s latest industrial motherboards, now supporting Intel Core 200S series processors (Bartlett Lake-S) through BIOS updates, are engineered to accelerate AI-driven applications at the edge. Offering flexible configuration with DDR5 and DDR4 memory across Mini-ITX, Micro-ATX, and ATX form factors, the series is tailored for smart manufacturing, machine vision, intelligent automation, and industrial AIoT and more.

https://asrockind.com

> Read More

About Smart Cities

This news story is brought to you by smartcitieselectronics.com, the specialist site dedicated to delivering information about what’s new in the Smart City Electronics industry, with daily news updates, new products and industry news. To stay up-to-date, register to receive our weekly newsletters and keep yourself informed on the latest technology news and new products from around the globe. Simply click this link to register here: Smart Cities Registration