I’ve already gone through the specifications, an unboxing, and a teardown of the GEEKOM GT1 Mega AI mini PC powered by an Intel Core Ultra 9 185H “Meteor Lake” processor with 32GB DDR5 RAM and a 2TB M.2 NVMe SSD in the first part of the review. I’ve now had more time to play with the GEEKOM GT1 Mega, so I will report my experience with Windows 11 Pro operating system in the second part of the review testing features, running benchmarks including an AI benchmark, evaluating networking and storage performance, testing the thermal design while under stress, and taking measurements for fan noise and power consumption. Software overview and features testing Going to the System->About window in the Settings confirms we have a GT1-Mega mini PC powered by a 2.50 GHz (base frequency) Intel Core Ultra 9 185H processor with 32GB of RAM running Windows 11 Pro 23H2 build [...]
iBASE Technology has unveiled the IB962 3.5-inch single-board computer (SBC) powered by the Intel Core Ultra 5/7 100 Series processor family (formerly Meteor Lake U/H) and delivering an “optimal balance of performance and power efficiency.” The IB962 3.5-inch SBC mostly follows the IB961 board‘s design, offering more powerful SoC options, more system memory, and an HDMI 2.0 display interface supporting 4K resolution at 60Hz. It provides dual Gigabit Ethernet networking, three USB 3.2 ports (two via pin headers), three USB 2.0 ports, and three M.2 slots (M-Key + E-Key + B-Key) for networking, storage, and cellular communication. While IB962 is not a single-board computer in the strictest sense since it uses SODIMM sockets, it is still quite compact and useful for portable or space-constrained applications. It is intended for industrial automation, manufacturing, retail, entertainment, healthcare, and various edge AI applications. It is a larger alternative to AAEON’s PICO-MTU4 Pico-ITX SBC [...]
Renesas has expanded its Arm Cortex-M85-based RA8 series to include two new entry-level MCUs, RA8E1 and RA8E2. The new microcontrollers perform similarly to other Arm Cortex-M85-based microcontrollers and are offered at a lower price point thanks to a streamlined feature set to reduce costs. The RA8E1 and RA8E2 groups of microcontrollers will be useful for high-volume applications in industrial and home automation, healthcare, office equipment, and consumer products. We have previously covered other models in Renesas’ RA8 series, including the Renesas RA8T1, RA8D1, and the very first Cortex-M85 microcontroller, the Renesas RA8M1. The RA8 series consists of microcontrollers based on the Arm Cortex-M85 core with a performance of 6.39 CoreMark/MHz and Arm’s M-Profile Vector Extension which provides a 4x performance boost for DSP and ML workloads while maintaining low power consumption. The RA8E1 and RA8E2 Cortex-M85 microcontroller groups differ mostly in clock speed, available internal and external memory, communication interfaces, [...]
Portwell has launched the PJAI-100-ON, a rugged and compact embedded system powered by the NVIDIA Jetson Orin Nano SOM and designed for demanding industrial environments. It operates reliably across a wide temperature range of -20°C to 60°C with a fanless, quiet design that requires minimal maintenance. Equipped with dual Gigabit Ethernet LAN ports and M.2 slots for wireless connectivity, the PJAI-100-ON is tailored for Edge AI applications. This system enhances quality control in manufacturing by using advanced optical inspection technology to detect defects, improving quality assurance and reducing production errors. Previously, we covered other Jetson Orin Nano embedded systems, including Aetina’s AIE-KO21, AIE-KO31, AIE-KN31, and AIE-KN41 and DFI X6-MTH-ORN fanless Edge AI Box computer, as well as Portwell’s WEBS-21J0-ASL another rugged embedded system built around an Intel Atom x7000RE Amstom Lake Nano-ITX motherboard and equipped with an Hailo-8 AI accelerator. Portwell PJAI-100-ON specifications: SoM – NVIDIA Jetson Orin Nano CPU [...]
Forlinx FET-MX95xx-C is a system-on-module (SoM) based on NXP i.MX 95 SoC with up to six Cortex-A55 cores, an Arm Cortex-M7 real-time core clocked at 800 MHz, an Arm Cortex-M33 “safety” core clocked at 333 MHz, and equipped with 8GB LPDDR4x and 64GB eMMC flash. The company also provides the feature-rich OK-MX95xx-C development board based on the i.MX 95 module with a wide range of interfaces such as dual GbE, a 10GbE SFP+ cage, terminal blocks with RS485 and CAN Bus interface, three USB Type-A ports, two PCIe slots, and more. Forlinx FET-MX95xx-C system-on-module Specifications: SoC – NXP i.MX 9596 CPU 6x Arm Cortex-A55 application cores clocked at 1.8 GHz (industrial) with 32K I-cache and D-cache, 64KB L2 cache, and 512KB L3 cache Arm Cortex-M7 real-time core clocked at 800 MHz Arm Cortex-M33 safety core clocked at 333 MHz GPU – Arm Mali-G310 V2 GPU for 2D/3D acceleration with support [...]
Day 4 of Giveaway Week 2024 will be for a development kit comprised of the Mixtile Core 3588E SoM based on Rockchip RK3588 octa-core Cortex-A76/A55 SoC and a Leetop A206 carrier board compatible with NVIDIA Jetson SO-DIMM modules. This development kit is mostly useful to evaluate the Mixtile Core 3588E system-on-module, but it can be used just like any Rockchip RK3588 SBC with the company providing a Ubuntu 22.04 Desktop image for the board. Since it relies on the same edge connector as the NVIDIA Jetson TX2 NX module, the Core 3588E module can also be connected to other compatible carrier boards. I reviewed the Mixtile Core 3588E development kit last December running Ubuntu 22.04. At the time, I found the Mixtile Core 3588E system-on-module performs well with the pre-loaded Ubuntu 22.04 image with a similar performance as on other Rockchip RK3588 hardware platforms. 3D graphics acceleration was working with [...]
In 2022, we covered Intel’s Xeon D (Ice Lake-D) processor family, which includes the D-2700 and D-1700 models. These processors are designed for software-defined networking and edge applications, providing data center-grade capabilities at the edge. Key features include integrated AI and crypto acceleration, built-in Ethernet, and support for Intel Time Coordinated Computing (TCC) and Time-Sensitive Networking (TSN) to ensure high reliability. Leveraging these advanced capabilities, companies like ADLINK Technology and Congatec have introduced COM-HPC server modules and COM Express Type 7 modules based on the new Xeon D lineup. Axiomtek’s CEM710 is a COM Express Type 7 Basic Module powered by Intel Xeon D-1700 Series processors, designed for edge computing, industrial automation, and data-intensive applications. It offers up to 10 cores, 100Gbps Ethernet, and 64GB of DDR4 ECC memory, along with PCIe Gen3 slots for expandability. With rugged features and TPM 2.0 for security, it ensures reliable performance in demanding [...]
Andes’ Voyager Development Platform is a micro-ATX motherboard based on the company’s QiLai SoC with four AX45MP 64-bit RISC-V cores and an NX27V vector processor optimized for AI workloads. The Voyager board supports up to 16GB of DDR4 via a UDIMM socket, features 16Mbit SPI Flash for boot code and an SD card socket, includes three PCIe Gen4 slots to connect external devices such as GPU card and SSD, as well as a JTAG debugger and USB to UART bridge for debugging during Linux software development. Andes QiLai RISC-V AI SoC Before looking at the board itself, let’s check the Andes QiLai SoC specifications: CPU cores Andes AX45MP-4C cluster 4x RV64GC 8-stage superscalar processors clocked at 1.6 GHz (worse) / 2.1 GHz (typ.) Support of MESI cache coherence protocol with Coherence Manager 32KB L1 Instruction and Date cache 2MB L2 cache I/O Coherence Port (IOCP) – Synchronous AXI4 (256 bits [...]
The M5Stack Module LLM is yet another box-shaped device from the company that provides artificially intelligent control without internet access. It is described as an “integrated offline Large Language Model (LLM) inference module” which can be used to implement local LLM-based solutions in smart homes, voice assistants, and industrial control. Module LLM is powered by the AX630C SoC, equipped with 4GB LPDDR4 memory, 32GB storage, and a 3.2 TOPS (INT8) or 12.8 TOPS (INT4) NPU. M5Stack says the main chip has an average runtime power consumption of 1.5W, making it suitable for long-term operation. It has a built-in microphone, speaker, microSD card slot, and USB OTG. The USB port can connect peripherals such as cameras and debuggers, and the microSD card slot supports cold and hot firmware updates. The M5Stack Module LLM joins the list of other offline, on-device LLM-based solutions, such as the SenseCAP Watcher, Useful Sensors’ AI in [...]
AAEON UP Squared Pro 710H is a 4×4-inch SBC and developer board designed for machine vision applications with an Intel Processor N97 or Core i3-N305 Alder Lake-N CPU and a 26 TOPS Hailo-8 Edge AI processor. The single board computer supports up to two MIPI CSI cameras via a 61-pin connector and offers expansion capabilities through a 40-pin Raspberry Pi-compatible header and three M.2 sockets for 5G, Wi-Fi, and NVMe storage support. While it’s only designed to operate in the 0 to 60°C temperature range, it does have industrial features such as 12V to 36V DC input and two RS-232/422/485 connectors. UP Squared Pro 710H specifications: Alder Lake N-series SoC (one or the other) Intel Processor N97 quad-core processor up to 3.6 GHz with 6MB cache, 24EU Intel UHD Graphics Gen 12 @ 1.2 GHz; TDP: 12W Intel Core i3-N305 octa-core processor up to 3.8 GHz with 6MB cache, 32EU Intel [...]
STMicroelectronics (STMicro) has announced the ST1VAFE3BX biosensor, a highly integrated chip that combines “cardio and neurological sensing with motion tracking and embedded AI functionalities,” and is targeted at healthcare and fitness wearables. The ST1VAFE3BX biosensor chip integrates biopotential inputs with an accelerometer and a machine learning core for reduced power consumption. It features a complete vertical analog front end (vAFE) for simplified detection of vital signs in electrocardiography, electroencephalography, and other healthcare monitoring applications. Inertial sensing by the accelerometer is synchronized with biopotential sensing “to infer any link between measured signals and physical activity.” The integrated machine-learning core (MLC) and finite state machine (FSM) allow for ultra-low-power and accurate biopotential input recognition at a degree of responsiveness that is reportedly faster than the current industry standard. The ST1VAFE3BX is aimed at applications in healthcare and fitness wearables such as smartwatches, smart patches, sports bands, connected rings, and smart glasses. It [...]
GEEKOM GT1 Mega is an AI mini PC based on the latest Intel Core Ultra 5/7/9 Meteor Lake processor. We received a review sample from the company with an Intel Core Ultra 9 185H SoC, 32GB DDR5 memory, and a 2TB NVMe SSD as well as an Intel BE200 WiFi 7 and Bluetooth 5.4 module. As usual, we’ll do a three-part review starting with an article with the specifications, an unboxing, and a teardown followed by a thorough review with the pre-installed Windows 11 Pro operating system, and finally with the latest Ubuntu 24.04.1 OS. GEEKOM GT1 Mega specifications GEEKOM provides the following specifications for the GT1 Mega mini PC and motherboard: Meteor Lake SoC (one or the other) Intel Core Ultra 9 185H 16-core (6P+8E+2LPE) processor @ 2.5 / 3.8 / 5.1 GHz (Turbo) with 24MB cache, 8x Xe-cores Intel Arc graphics @ 2.35 GHz, Intel AI Boost NPU; [...]
Kontron has launched its new Amston Lake Pico-ITX SBC, the 2.5”-SBC-AML/ADN series, optimized for low-power, real-time computing in industrial automation, smart cities, energy, retail, and healthcare applications. This compact SBC series supports Intel’s energy-efficient processors, including the Intel Atom x7000E Series, Intel Core i3 N-Series, Intel N-Series (Alder Lake-N), and the Intel Atom x7000RE Series (Amston Lake). With Intel UHD Graphics (Gen12) featuring up to 32 execution units, these Amston Lake Pico-ITX SBCs enhance graphics processing and support concurrent AI inference. Powered by Intel Deep Learning Boost, they achieve faster AI inference, which is ideal for efficient edge computing, while meeting the stringent demands of real-time, low-latency applications. Previously, we covered several Pico-ITX SBCs, including the AAEON PICO-ASL4, Jetway JPIC-ADN1, AAEON PICO-IMX8PL, and AAEON RICO-MX8P, among others. Feel free to explore these reviews if you’re interested in similar SBCs. Kontron 2.5”-SBC-AML/ADN specifications: CPU Standard Intel Atom x7211RE dual-core processor up [...]
Seeed Studio has recently launched the OpenUC2 10x AI portable microscope built around the XIAO ESP32-S3 Sense module. Designed for educational, environmental research, health monitoring, and prototyping applications this microscope features an OV2640 camera with a 10x magnification with precise motorized focusing, high-resolution imaging, and real-time TinyML processing for image handling. The microscope is modular and open-source making it easy to customize and expand its features using 3D-printed parts, motorized stages, and additional sensors. It supports Wi-Fi connectivity with a durable body, uses USB-C for power and swappable objectives make it usable in various applications. Previously we have written about similar portable microscopes like the ioLight microscope and the KoPa W5 Wi-Fi Microscope, and Jean-Luc also tested a cheap USB microscope to read part number of components. Feel free to check those out if you are looking for a cheap microscope. OpenUC2 10x specifications: Wireless MCU – Espressif Systems ESP32-S3 CPU [...]
Microchip recently unveiled the PIC64HX1000 64-bit RISC-V AI microprocessor (MPU) family designed for mission-critical intelligent edge applications in the aerospace, defense, industrial, and medical sectors thanks to a quantum-resistant design. These new MPUs feature eight SiFive’s Intelligence X280 cores, each clocked at 1 GHz. The MPUs are engineered with a decoupled vector pipeline enabling 512-bit operations enabling the PIC64HX1000 to achieve up to 2 TOPS for AI/ML processing and come equipped with integrated Time-Sensitive Networking (TSN) Ethernet connectivity. This new microprocessor includes a dedicated system controller for runtime monitoring and fault management, WorldGuard architecture for workload isolation, and post-quantum defense-grade cryptography, which includes the NIST-standardized FIPS 203 and FIPS 204 cryptographic algorithms, ensuring protection against future quantum computing threats. PIC64HX1000 64-bit AI MPU specification MPU variants PIC64HX1000 – IN (Industrial) PIC64HX1000 – AV (Aviation) PIC64HX1000 – MI (Military) CPU – 8x 64-bit RISC-V cores (SiFive X280), up to 1 GHz, [...]
Portwell’s WEBS-21J0-ASL is a fanless, rugged embedded system with a Nano-ITX motherboard powered by an Intel Atom x7000RE Series processor and optimized for industrial IoT applications. It features two 2.5Gbps Ethernet ports, dual 4K display support (DP/HDMI), and a serial port as well as M.2 E and B Key sockets for wireless and storage expansion. Designed for harsh environments, the WEBS-21J0-ASL operates in temperatures ranging from -20°C to 60°C. The system is equipped with DDR4 SO-DIMM memory (up to 16GB and 10Gbps USB 3.2 Gen 2 ports. It has passed vibration and shock tests for durability in rugged applications making it suitable for a wide range of industrial and automation systems. The WEBS-21J0-ASL also supports optional integration with the Hailo-8 AI acceleration module delivering up to 26 TOPS of AI performance for applications requiring advanced AI capabilities. This makes it an ideal solution for robotics, medical equipment, outdoor gateways, and [...]
Waveshare’s Jetson Orin NX Development Kit is an AI edge computing platform tailored for robotics and AI-driven applications built around the NVIDIA Jetson Orin NX module 8GB/16GB delivering up to 70 TOPS and 100 TOPS of AI performance, respectively. It features the JETSON-ORIN-IO-BASE base board, which provides essential interfaces like an M.2 socket, DisplayPort (DP) video output, and USB ports, making it easy for users to connect multiple sensors and peripherals for high-performance AI tasks. The kit includes a 128GB NVMe SSD for high-speed storage and comes pre-installed with the AW-CB375NF wireless network card, which supports dual-band Wi-Fi 5 and Bluetooth 5.0. The Jetson Orin NX module operates within a configurable power range of 10W to 25W, delivering up to 5x the performance of the Jetson Xavier NX module. This development kit is suitable for large-scale AI projects requiring multiple concurrent inference pipelines and high-speed data processing. Previously, we covered [...]
The AI HAT+ features the same best-in-class Hailo AI accelerator technology as our AI Kit, but now with a choice of two performance options: the 13 TOPS (tera-operations per second) model, priced at $70 and featuring the same Hailo-8L accelerator as the AI Kit, and the more powerful 26 TOPS model at $110, equipped with the Hailo-8 accelerator.
Designed to conform to our HAT+ specification, the AI HAT+ automatically switches to PCIe Gen 3.0 mode to maximise the full 26 TOPS of compute power available in the Hailo-8 accelerator.
Unlike the AI Kit, which utilises an M.2 connector, the Hailo accelerator chip is directly integrated onto the main PCB. This change not only simplifies setup but also offers improved thermal dissipation, allowing the AI HAT+ to handle demanding AI workloads more efficiently.
What can you do with the 26 TOPS model over the 13 TOPS model? The same, but more… You can run more sophisticated neural networks in real time, achieving better inference performance. The 26 TOPS model also allows you to run multiple networks simultaneously at high frame rates. For instance, you can perform object detection, pose estimation, and subject segmentation simultaneously on a live camera feed using the 26 TOPS AI HAT+:
Both versions of the AI HAT+ are fully backward compatible with the AI Kit. Our existing Hailo accelerator integration in the camera software stack works in exactly the same way with the AI HAT+. Any neural network model compiled for the Hailo-8L will run smoothly on the Hailo-8; while models specifically built for the Hailo-8 may not work on the Hailo-8L, alternative versions with lower performance are generally available, ensuring flexibility across different use cases.
After an exciting few months of AI product releases, we now offer an extensive range of options for running inferencing workloads on Raspberry Pi. Many such workloads – particularly those that are sparse, quantised, or intermittent – run natively on Raspberry Pi platforms; for more demanding workloads, we aim to be the best possible embedded host for accelerator hardware such as our AI Camera and today’s new Raspberry Pi AI HAT+. We are eager to discover what you make with it.
The Raspberry Pi AI HAT+ is a PCIe expansion board for the Raspberry Pi 5 with either a 13 TOPS Hailo-8L or 26 TOPS Hailo-8 AI accelerator. You may remember the Raspberry Pi AI Kit was introduced last June with an official M.2 Key M HAT+ and a 13 TOPS Hailo-8L M.2 AI accelerator module, The new Raspberry Pi AI HAT+ is quite similar except the chip is soldered on the expansion board and offered with either Hailo-8L or the more powerful Hailo-8 variant. Raspberry Pi AI HAT+ specifications: Supported SBC – Raspberry Pi 5 AI accelerator Hailo-8L AI accelerator with up to 13 TOPS of performance Hailo-8 AI accelerator with up to 26 TOPS of performance Host Interface – PCIe Gen3 interface 16mm stacking GPIO header PCIe FPC cable Spacers and screws enabling fitting on Raspberry Pi 5 with Raspberry Pi Active Cooler Dimensions – 65 x 56.5 mm [...]
MentorPi is a ROS2-compatible robot car powered by the Raspberry Pi 5, designed for AI-driven robotics and Python programming. It offers two chassis options: MentorPi-M1, which features a Mecanum-wheel chassis, and MentorPi-A1, equipped with an Ackermann chassis. Both variants come with high-performance components such as closed-loop encoder motors, STL-19P TOF lidar, 3D depth cameras, and high-torque servos. These enable precise navigation, SLAM mapping, path planning, and dynamic obstacle avoidance, making MentorPi an ideal platform for robotics tasks. The system utilizes a dual-controller architecture to optimize performance. The Raspberry Pi 5 handles AI vision processing and strategic functions, while Hiwonder’s RRC Lite expansion board manages motion control and sensor data processing. This task distribution enhances efficiency in machine vision, AI-powered navigation, and robotic control, allowing MentorPi to tackle complex AI and vision-based applications with ease. MentorPi also supports advanced features like 3D visual mapping and YOLOv5-based object detection for recognizing road [...]
Login
