X-SENSE SC07-WX WiFi Smoke & CO Alarm Review – Your Smart Home Guardian for Smoke and CO

The X-Sense SC07-WX is a battery-powered smoke and carbon monoxide detector that combines Wi-Fi connectivity with a smartphone app for remote monitoring and alerts.  The sensor uses a photoelectric smoke sensor and an advanced electrochemical carbon monoxide sensor for fast and accurate detection.

The device has its own Android and iOS application and directly connects to the smartphone so there’s no need for a docking station of some sort. The application also will send you push notifications in case it detects any smoke in the area, other than that it also sends you a notification when there’s low battery, faults, and others.

The overall life cycle of the product is about 10 years and is powered by a single CR123A 3V Lithium Battery. Let me be obvious – the device has a life cycle of 10 years, meaning some internal counters inside the device will make the device unusable after 10 years! like it or not but is per the law in major countries. But the battery inside it will not last for that long and you may have to change it periodically.

There is also a small screen in front of the module that helps you stay informed by indicating the carbon monoxide level in parts per million (PPM), showing the battery life, and displaying the device’s status on the go, there’s also a LED bar just under the display that also is an indicator it flashes up green yellow for fault and alarm.

In the package, you’ll find everything you need to set up and utilize the SC07-WX smoke and carbon monoxide detector efficiently. Alongside the detector itself, there’s a mounting base for easy installation. There also be a user manual that provides clear instructions for setting up a warranty card, and carbon monoxide reference stickers. These stickers serve as helpful reminders to stay alert to potential carbon monoxide dangers. To mount it on your rooftop or wall you just need to mount the back plate and the module just gets screwed into the plate.

Note! Before installing the device you need to make sure that you are in your router’s wifi range, the module gets connected to your wifi through your 2.4GHz link so you also need to make sure that your 2.4GHz is enabled.

X-SENSE SC07-WX Smoke and Carbon Monoxide Allerm Specifications

  • Model: SC07-WX
  • Manufacturer: X-Sense
  • Transmission Range: Covers up to 170 feet (50 meters)
  • Smoke Detection: Utilizes photoelectric technology
  • Carbon Monoxide Detection: Employs electromechanical sensors
  • Detection range:
    • 70ppm for 60-240 minutes
    • 150ppm for 10-50 minutes
    • 400ppm for 4-15 minutes
  • Wi-Fi: Operates on 2.4GHz 802.11 b/g/n network (not compatible with 5GHz network)
  • Batteries: 1 CR123A battery required (included)
  • Size: 1-Pack
  • Style: Wi-Fi Connected Model
  • Shape: Circular
  • Power Source: Battery Powered
  • Installation Method: Screw-In
  • Mounting Type: Protruding
  • Usage: Inside
  • Batteries Included?: Yes
  • Batteries Required?: Yes
  • Item Weight: 1.06 pounds
  • Product Dimensions: 5.7 x 5.7 x 2 inches

Video

In the package, you’ll find everything you need to set up and utilize the SC07-WX smoke and carbon monoxide detector efficiently. Alongside the detector itself, there’s a mounting base for easy installation. There also be a user manual that provides clear instructions for setting up a warranty card, and carbon monoxide reference stickers. These stickers serve as helpful reminders to stay alert to potential carbon monoxide dangers. To mount it on your rooftop or wall you just need to mount the back plate and the module just gets screwed into the plate.

If you like this device and want to purchase one for your home, apartment, dwelling, hovel, or shack you can get them on both Amazon and X-Sense’s website. On Amazon, you can buy a single pack for $44.99 or opt for a three-pack at a discounted price of $129.99. Similarly, on X-Sense’s website, the pricing remains consistent.

ARM Unveils Ethos-U85 NPU and Corstone-320 Platform for Enhanced Edge AI

Arm-Ethos-U85

Arm Targets the AIoT with High-Performance Ethos-U85 NPU and Corstone-320 Platform New accelerator boasts four times the peak performance and a 20 percent power efficiency boost over its predecessors. Arm has announced its latest designs for the Artificial Intelligence of Things (AIoT) and edge artificial intelligence (edge AI): the new, faster Ethos-U85 neural processing unit (NPU) and a reference design platform, the Corstone-320, which combines it with an Arm Cortex-M85 CPU and a Mali-C55 image signal processor.

“Our family of Ethos-U NPUs were the world’s first embedded AI accelerators and they’re already in silicon from leading players,” Arm’s Paul Williamson told us during a pre-launch briefing on the company’s latest designs.

“[The] Ethos-U85 delivers a 4x performance uplift for high-performance edge AI applications. Along with the 4x performance improvement, the U85 also delivers 20 percent higher power efficiency over the previous generation, is scalable from 128 to 2,048 MAC units, and delivers four TOPS [Tera-Operations Per Second at INT8 precision] at the highest performance configuration.”

The Ethos-U85 not only offers increased power but also introduces support for transformer networks, which allows for quicker customization and optimization of edge AI applications. This feature enables transformer-based models to be easily adapted to various tasks compared to convolutional networks, making efficient use of hardware resources and suitable for deployment on edge devices with limited computing capabilities. Additionally, the Ethos-U85 maintains backward compatibility with earlier Ethos-U NPU models, ensuring seamless integration with existing toolchains. Meanwhile, the Corstone-320 reference design platform combines the Ethos-U85 NPU with Arm’s Cortex-M85 CPU core and Mali-C55 image signal processor to provide a ready-to-use solution for edge AI development, including virtual hardware support before silicon availability.

Silicon isn't expected until 2025, but the Corstone-320 reference platform includes virtualization options
Silicon isn’t expected until 2025, but the Corstone-320 reference platform includes virtualization options

However, the Ethos-U85 is not restricted to Arm’s Cortex-M microcontroller-class components alone. Williamson indicated that the NPU is also compatible with Cortex-A application-class cores, allowing for acceleration of machine learning and artificial intelligence tasks in various applications such as robotics, industrial machine vision, and wearables.

The Ethos-U85 system configurations

Arm has not yet shared detailed benchmarks or power measurements for the Ethos-U85, which will vary by configuration, but Williamson has positioned it as “a milliwatts-level power envelope” design. The company has confirmed it is licensing the NPU to “early adopters” including Alif and Infineon now, with Williamson telling us that silicon is expected to land on the market sometime in 2025.

More information on the Ethos-U85 is available on the Arm blog.

Introducing SenseCAP Watcher: Seeed Studio’s AI-Powered Smart Monitoring Solution

Seeed Studio's SenseCAP Watcher Combines Local TinyML with Cloud LLMs to Deliver Smart Monitoring

Seeed Studio’s SenseCAP Watcher Combines Local TinyML with Cloud LLMs to Deliver Smart Monitoring

Seeed Studio’s SenseCAP Watcher is a cutting-edge device that merges the power of local TinyML (Tiny Machine Learning) with cloud-based Large Language Models (LLMs) to provide intelligent monitoring solutions. This innovative gadget acts as a “physical AI agent” designed to enhance the functionality of various spaces, leveraging advanced AI technology and IoT connectivity.

The SenseCAP Watcher integrates locally executed TinyML models with cloud-based LLMs, allowing it to analyze data efficiently while benefiting from the vast computational resources available in the cloud. By combining these two approaches, the device can deliver smart monitoring capabilities that enable users to gather insights and make informed decisions in real-time.

Seeed Studio has opened applications for alpha testers, signaling its commitment to refining and enhancing the SenseCAP Watcher before its official launch. With this initiative, the company aims to democratize access to modern artificial intelligence, making it accessible and applicable in diverse environments and settings.

“We are incredibly excited to launch the SenseCAP Watcher at Embedded World 2024,”

says Seeed Studio Eric Pan, of the company’s latest hardware release.

“With the SenseCAP Watcher, we envision a more effortless way of interacting with and managing the physical world. It pushes the boundaries of what a combination of tinyML at the edge and LLMs can enable while retaining an open source and extensible design.”

Designed for versatile use, the SenseCAP Watcher offers both wall mount and desktop operation options. Its feature set includes a 1.46″ circular color touchscreen display, a front-facing camera and microphone, a speaker, programmable RGB light, wheel interface, and two USB Type-C connectors for power and charging its internal battery. Additionally, it boasts a Grove expansion slot for attaching external add-on hardware, enhancing its functionality and customization possibilities. This gadget serves as a key component of Seeed’s  SenseCraft, toolkit, which seamlessly integrates on-device tinyML models with remote large language models. These models can run either on remote cloud servers or on local edge devices, such as Seeed’s Edge AI Box reComputer devices based on the NVIDIA Jetson system-on-module family.

Seeed Studio’s SenseCAP

While Seeed showed the device at Embedded World 2024 in Nuremberg, it has yet to announce pricing and availability — but is taking applications for alpha testers on its website, where interested parties can also interact with a simulated version of the SenseCAP Watcher dubbed “Nobody.”

ArmSom Sige7 Review- A Rockchip RK3588 SBC with Dual 2.5GbE Ethernet, NVMe Storage, and Triple Display Output

Introduction

The ArmSoM Sige7 is a compact single-board computer (SBC) powered by the RK3588, an octa-core 64-bit SoC clocked at 2.4 GHz. The SoC features a 6TOPS NPU for AI tasks and supports up to 32GB of LPDDR4x RAM and 128GB of eMMC flash storage. It includes an M.2 2280 socket for NVMe SSDs. It offers three display interfaces (HDMI, USB-C, MIPI DSI) and two camera connectors. Connectivity options include dual 2.5GbE, WiFi 6, and Bluetooth 5.2, plus multiple USB ports and a 40-pin GPIO header for expansion.

The SoC features the ARM Mali-G610 MP4 GPU clocked at 1GHz, which delivers high-performance graphic capabilities so much that it can play or stream 4K very smoothly. This GPU also features enhanced 3D rendering which allows for lag-free gameplay in major open-world online games, console games, and even some AAA titles, this feature makes this device suitable for a broad spectrum of gaming enthusiasts looking for a powerful, compact computing solution.

RK3588 Features:

  • 8K video encoding/decoding support
  • Abundant interfaces support
  • Dual 2.5G Ethernet support
  • WiFi 6 & BT 5 support
  • Multiple video outputs support
  • Multiple operating systems support

Triple Display Interface

The most interesting feature for me is the triple display interface, When you think about applications where you will need a triple display like video editing graphic design 3d modeling, the board is not powerful enough to run that application. But it will shine when it comes to Trading, you can just use this device as a multiple equities monitoring device.

Specifications

  • CPU: Quad-Core [email protected]+Quad-CoreCortex-A55@ 1.8GHz,8nm process
    • 6 TOPS@INT8(3 NPU core)
  • GPU&VPU: GPU Mali-G610 MP4 (4x256KB L2 Cache)
    • Decode: 8K@60fps
    • Encode:8K@30fps H.265 / H.264
  • Storage: 8GB LPDDR4x + 64bit 64GB eMMC 5.1
  • Interface: 2x 2.5G Ethernet
    • Onboard IEEE 802.11a/b/g/n/ac/ax WIFI6 and BT5 (AP6275P)
    • 1x HDMI 2.1, supports 8K@60fps
    • 1x DP 1.4 up to 8192×4320@30Hz
    • 1x MIPI DSI up to 4K@60Hz
    • 2x 2-lane MIPI CSI, up to 2.5Gbps per lane

Neural Processing Unit

The RK3588 processor includes a powerful Neural Processing Unit (NPU) capable of performing up to 6 trillion operations per second, enhancing its ability to handle AI tasks efficiently. This NPU supports well-known deep learning frameworks like TensorFlow, PyTorch, and MxNET, broadening its application in various AI fields. With this capability, the RK3588 is optimized for AI applications, offering improved performance in tasks such as image and voice recognition, making it a versatile choice for AI-driven projects.

8K UHD support

Among other features, this board also has advanced video encoding and decoding capabilities with 8K UHD support. This allows for ultra-high-definition video playback and recording, enhancing applications in digital signage, entertainment, and gaming with its capacity to deliver sharp, detailed visuals. It’s designed to create immersive display experiences across different settings, significantly enhancing user engagement.

ArmSoM Sige7 Supported Linux Distributios

OS Support

The board supports Debian 11 and Android 12, based on Linux 5.10, and is compatible with the Buildroot build system. Additionally, third-party Armbian images (Ubuntu 20.04, Ubuntu 22.04) and Kylin OS are available. The wiki currently has direct links to these images.

Update! In a recent update, the company tells electronics-lab.com that they have “successfully integrated Linux into the mainline” meaning the developers have ensured that their ArmSoM-Sige7 device will be compatible with the widest range of standard Linux distributions and software. This avoids the need to maintain custom versions of Linux. Mainline support ensures that the Sige7 will benefit from future bug fixes, security patches, and new features added to the Linux kernel over time.

Connectivity

The board features the AP6275P Wi-Fi BT module that offers advanced wireless capabilities, including support for the latest 802.11a/b/g/n/ac/ax standards for Wi-Fi and BT including 2×2 MIMO technology for better signal quality and reliability. All and all this module is designed for robust wireless connectivity, enabling high-speed internet access and efficient data transmission across a wide range of applications and devices.

Resources and Development

The ArmSom Sige7, also known as the Banana Pi BPi-M7, offers detailed resources for developers and users, this includes links to kernel and u-boot sources hosted on the ArmSom GitHub account, highlighting the collaborative and open-source nature of the project. Other than that, ArmSom provides an array of in-development documentation, firmware images, and essential development tools, all accessible via Google Drive.

Multiple Use Cases

The board has a lot of use cases, which makes it perfect for a bunch of projects. You can use it as a mini-computer whether you’re at home, at work, or on the move. It’s also great for setting up a tiny server to manage web services or files. For those into AI and robotics, this board has the chops to handle complex projects. Plus, with edge computing capabilities, it can process data super fast, cutting down on lag. It’s a little powerhouse for any tech project you’ve got in mind.

Packaging

Board Outline

The company also offers a handy hardware description for the board, highlighting all the key components. This guide makes it easier for both developers and users to get started with the board smoothly and effectively.

More information & Purchase

The Sige7 documentation includes a product introduction, a detailed manual, performance benchmarks, and access to a user forum and Discord. This comprehensive set of resources is designed to help both developers and users quickly get up to speed with the board, for more information on the board you can check out the ArmSom Sige7 docs page or other store page & Aliexpress.

Exciting Update!  The company also confirms the launch of  ArmSoM’s RK3588 AI Module (AIM7) on Crowdsupply. Stay tuned for updates and your chance to support this project!

Giveaway is coming next week. Stay tuned!

CrowVision 11.6” Capacitive Touch Display Review – Compatible with Raspberry Pi, BeagleBone, Jetson Nano, and More

The CrowVision is an 11.6IPS Capacitive Touchscreen Display module specially designed to be attached to most single-board computers(SBC) with adjustable mounting holes in the back.

The display features a resolution of 1366×768, so for an 11.6” it translates to 135ppi. It offers a wide viewing angle of 178 degrees. Connectivity options include a USB micro socket for touch functionality, with a mini HDMI port as the display input.

Unboxing

Once the box arrives at your doorstep you will find a 12V/2A power supply, USB Type-C and micro USB cords, HDMI to mini HDMI and micro HDMI to mini HDMI connectors, two cable organization “ribbons,” a screwdriver, an on-screen display (OSD) control panel with five buttons, and a guide for users all available in the box.

Package List

  • 1x 11.6 inch capacitive Touch Screen
  • 1x USB-A to USB-C cable
  • 1x USB-A to Micro B cable
  • 1x HD to Mini HD cable
  • 1x Micro HD to Mini HD cable
  • 1x OSD Control Board
  • 1x Power Adapter
  • 1x Screwdriver
  • 2x Ribbon
  • 1x User Manual

Display Driver

The company provides a detailed description of the control board, for you to get started easily, in the same image you can see other brass mounting holes through which you can attach supported SBC just with three screws. Speaking of the display, after a close inspection I found out that the display is built around an RTD2556 display controller which is quite a popular controller and used often as a display driver. While the RP4 was booting we saw some artifacts on the small fonts on the screen, and we thought this was due to the low resolution of the device.

Specifications

  • Size: 11.6 Inch
  • Touch Type: 5-point Capacitive Touch
  • Resolution: 1366*768
  • Color Depth: 16M
  • Viewing angle: 178°Wide Viewing Angle
  • Display Type: IPS Panel
  • Screen Type: TFT-LCD
  • External power supply: 12V-2A
  • Digital input: HDMI-compatible interface
  • Interface: 1xKeypad interface, 1x Power supply 5V output, 1x Mini HD interface, 1xTouch interface, 1x Speaker interface, 1x Headphone socket,1x Power supply 12V input
  • CompatibilitySystem: Raspbian, Ubuntu, Windows, Android, MAC OS, and Chrome OS,etc.
  • Active Area: 256.13*144mm(L*W)
  • Dimension Size: 290.8*184.2mm(L*W)

The display will not come with an enclosed instead on the sides of the display you will have 3M adhesive tape which will be useful if you are making your custom enclosure for the board.

Features

  • 11.6-inch high-resolution screen with 1366*768 resolution, IPS panel, and 178° wide viewing angle provides a better visual experience
  • Unique rear fixing structure with sliding fixing pillars, compatible with most single-board computer models, easy to assemble
  • Wide compatibility, compatible with multiple operating systems (Raspbian, Ubuntu, Windows, Android, MAC OS, and Chrome OS)
  • Supports audio, video, and capacitive touch, plug and play
  • Integrates a variety of peripheral interfaces (such as speakers, headphones, keypads, touchscreens) and onboard OSD adjustment keys
  • The mainboard is equipped with power conversion function of output 5V/3A, not need to separately connect an external power supply for the single-board computer.

Support various SBCs

For convenience, the company also provides a list of supported SBCs along with mounting instructions, once you secure your SBC in the back of the display you need to connect, a mini HDMI to micro HDMI cable for video output, and a USB to USB-C cable for power. One downside of this setup is that as this board offers versatility there will always be a cable mess of cable when using this board. You can see the mess in the following photo. This could be managed with smaller or specially designed cables, but we use generic ones.

Power Delivery

Although it’s theoretically possible to connect various SBCs within the supported dimensions, one thing to note is the power output from its control board. It maxes out at 5V/3A, which, for the most part, does the job. But, if you’re planning on hooking up some of the more power-hungry models, you might find it falling short. It’s not that important but something to keep in mind. I had to dig out an extra power source for one of my setups, but once that was sorted, it was smooth sailing.

Plug n Play

Elecrow’s done their homework, testing many computing devices to ensure they play nice with the CrowVision. It’s reassuring to see a company not just throw tech specs around but back them up with real-world testing. For the devices that are directly supported, it’s pretty much a plug-and-play experience, which is fantastic.

Conclusions

The board is good but there are two small problems with this display module, the first one is that they do not provide an enclosure for the display but for those who want to get an enclosure quickly they provide .OBJ files for easy going. You can always print the enclosure or you can give it to the fab house for them to print. The second is the low resolution of the display which makes the smaller elements not to seen clearly.

In summary, if you’re into tinkering with SBCs or looking for a reliable control board for your projects, the CrowVision is worth considering.

More information

The company also provides a wiki page for further details but note thing to be careful of is its power output limitations with certain models, and you should be good to go.

Purchasing

If you want to purchase this cool and new little display module you can find the display module for $119.90 on Elecrow store. Happy tinkering!

TerraMaster Launches the Industry’s First 8-bay 10Gbps Hybrid Storage with an Early Bird 33% Discount

TerraMaster, a professional brand that focuses on providing innovative storage products for homes and businesses, has recently released the D8 Hybrid which is the industry’s first hybrid HDD NVMe SSD enclosure. It can hold 4 SATA HDDs/SSDs and 4 M.2 2280 NVMe SSDs, with a capacity of 24TB a drive, and 8TB per M.2 SSD, providing users with up to 128TB (24TB x 4 + 8TB x 4) of storage space, which is an excellent choice for storage expansion of Windows, Mac or Linux computers.

Combining the advantages of HDDs’ massive capacity and SSDs’ high-speed performance, and TerraMaster’s complimentary TPC Backupper software, you can easily schedule backups of important data to enhance data security.

D8 Hybrid Key Features

Hot and Cold Storage Combination

It is flexible to manually choose where to store your data based on type and frequency of use, thereby optimizing storage efficiency and access speed. Taking advantage of the large capacity of mechanical hard disks and the high-speed read and write performance of M.2 SSDs, hot data stored on high-speed SSDs can be quickly accessed, while cold data is cost-effectively stored on HDDs.

TPC Backup Solution

TerraMaster’s complimentary TPC Backupper simplifies Windows PC backups. Users can Schedule folder or disk partition backups to TerraMaster USB HDD storage or NAS servers effortlessly. Just connect, install, and set up your backup task. TPC Backupper supports both incremental and differential backup strategies and is compatible with Windows 8/8.1/10/11. For Mac users, you can seamlessly utilize the built-in Time Machine feature for backup purposes.

10Gbps Transfer Bandwidth

The D8 Hybrid hard drive enclosure adopts USB 3.2 Gen2 protocol for high-speed data transmission up to 10Gbps. With a single M.2 SSD (Samsung 980 NVMe M.s SSD), the read/write speed can reach up to 980MB/s. With 2 SSD’s in RAID 0, the read speed can reach 960MB/s, With 2 hard drives in RAID 0, the read/write speed can reach up to 521MB/s.

The 2+6 Innovation

Dive into the uniqueness of our 2+6 RAID mode! Slots 1 and 2 offer a playground for RAID 0/1/JBOD configurations, ensuring your data’s absolute security. Meanwhile, slots 3 and 4 and M.2 NVMe SSDs stand strong as independent guardians, catering to your every storage whim.

Unlock Infinite NAS Space

The D8 Hybrid is an ideal choice for expanding the NAS capacity of a home or small office. Simply connect the D8 Hybrid to the USB port of your NAS device, and you can easily expand the NAS storage space. You can even use the D8 Hybrid as an external DAS storage device for TerraMaster NAS. Both TOS 5 and the latest TOS 6 systems support creating new volumes using the four HDDs and four M.2 SSDs of the D8 Hybrid, and utilizing the full range of TerraMaster NAS storage features, such as setting up storage pools, various RAID modes, and enabling snapshots.

Push-Lock and Secure

Enjoy tool-free HDD tray swaps with Push-Lock auto-locking, preventing any accidental disconnections. Plus, slide in M.2 SSDs with specially designed covers and screws for a speedy 5-second install!

Stay Cool and Quiet

The D8 Hybrid features an intelligent cooling system, coupled with a super quiet design, which ensures your hard drives stay cool and your workspace stays serene. With automatic fan speed adjustment based on hard drive temperature and noise levels as low as 21dB, enjoy optimal performance without the extra noise.

The D8 Hybrid crowdfunding project has officially launched on Kickstarter with up to 33% off. For more details, please visit:

Follow TerraMaster on social media:

Arduino Introduces 4G Global Connectivity for Portenta in Mini-PCIe Form Factor

Arduino Pro has introduced a groundbreaking solution for IoT development with the launch of 4G global connectivity for its Portenta board, now available in a compact Mini-PCIe form factor. This innovation enables developers to swiftly create IoT devices with seamless global 4G cellular connectivity and GNSS capabilities.

Arduino made a significant announcement at Embedded World 2024, unveiling expansions to its esteemed Portenta line tailored for professional applications. These extensions integrate cellular connectivity, marking a notable advancement in the capabilities of Arduino’s Portenta platform. The Arduino Pro 4G module provides global cellular coverage and GNSS positioning in a mini-PCIe form factor. The company also unveiled a carrier board targeting its existing Portenta family and the new 4G module.

They introduced two new Arduino Pro 4G modules, each featuring a different system-on-module (SoM) design and adopting the mini-PCIe card form factor. One side of these modules features an edge connector, while the other side includes antenna connectors.

One of the 4G boards is tailored for the EMEA (Europe, Middle East, and Africa) region and incorporates a Quectel EC200A-EU modem. The other 4G board offers global coverage and employs a Quectel EG25-G modem, additionally providing global navigation satellite system (GNSS) positioning. The GNSS support encompasses GPS, GLONASS, BeiDou (COMPASS), Galileo, and QZSS, ensuring comprehensive positioning capabilities.
The Arduino Pro 4G modules support 4G, 3G, and 2G networks, making them versatile for various applications such as remote maintenance, fleet management, smart cities, and smart buildings. With their mini PCI-E card form factor, integration into devices is straightforward. Arduino’s carrier board simplifies prototyping, and the off-the-shelf connector facilitates easy deployment onto custom boards.

Arduino Pro 4G GNSS Module Global

The Portenta Mid Carrier combines a Portenta C33Portenta H7, or Portenta X8 development board with the Arduino Pro 4G module. Together, you can build devices that use Wi-Fi and cellular connectivity. The carrier board for the Arduino Pro 4G modules features numerous connectors and breakouts tailored for Portenta high-density expansion headers. Notably, it includes dedicated connectors for USB-A and Ethernet, enhancing connectivity options for users. It also features JTAG interfaces for debugging and the same connector styles as the Arduino GIGA for a display shield and camera.

All three products are now available from the Arduino Store — the Arduino Pro 4G Global (with GNSS) module at $65 and the EMEA module for €29. Neither comes with antennas, but Arduino offers Antenna Kits for each. The Portenta Mid Carrier costs $65.

Syntiant Unveils NDP250: A Breakthrough in Microwatt AI Processing for Always-On Applications

Syntiant's Third-Gen "Neural Decision" Chip

Syntiant, a leading provider of on-device deep learning solutions, has introduced its latest innovation: the NDP250 Neural Decision Processor. Boasting an impressive 30 giga-operations per second (GOPS) performance, the NDP250 sets a new standard for efficiency in AI processing. This third-generation chip represents a significant leap forward, offering five times the speed of its predecessor, the Core 2.

“Our NDP250 builds on two generations of neural network architectures to deliver 30 GOPS, making it our fastest, highest-performing chip yet,”

says Syntiant CEO Kurt Busch, who unveiled the chip at Embedded World.

Designed for always-on voice and vision applications, the NDP250 can power various tasks, from speech synthesis to image recognition. Its advanced architecture, including the innovative Core 3 engine, enables seamless multitasking while operating within a microwatt power envelope.

Busch further emphasizes the versatility and efficiency of the NDP250, stating, “Compatible with a host of architectures while running multiple different layers simultaneously at significantly less power than existing solutions, the NDP250 with our new Core 3 engine is the ideal real-time speech interface for large language models. It can bring powerful AI to battery-powered, always-on vision applications in automotive security, appliances, cameras, smart displays, and video doorbells.” He highlights the transformative impact of the NDP250, noting that applications that once demanded power measured in watts can now operate with power measured in microwatts.

The NDP250 adopts a familiar approach seen in the company’s previous “Neural Decision Processors,” but with the introduction of the new third-generation Syntiant Core 3 architecture. This upgrade brings significant enhancements, with the company claiming a fivefold increase in performance compared to its second-generation counterparts, all without increasing power requirements. The chip’s architecture includes an Arm Cortex-M0 processor alongside Syntiant’s Core 3 and a HiFi3 programmable digital signal processor (DSP). Additionally, it features a dual 11-wire direct image interface, dual PDM digital microphone interface, I2S with PCM support, and quad-SPI and dual-I2C buses, complemented by 58 general-purpose input/output (GPIO) pins.

According to Syntiant, the NDP250 excels in handling multiple heterogeneous networks concurrently, supporting over six million total neural parameters when operating in eight-bit precision mode. It can accommodate various neural network architectures, including 1D, 2D, and depthwise convolutional neural networks (CNNs), fully-connected networks, and recurrent neural networks such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. Besides real-time speech recognition and voice synthesis tasks, the chip is capable of efficiently running always-on image recognition models while consuming less than 30mW of power.

Syntiant is showcasing the NDP250 at its booth during Embedded World 2024, located at Hall 2 Booth #2-238. The company has confirmed that the chip is currently available for sampling; however, it has not provided a timeline for general availability.

More information on Syntiant’s chips is available on the company’s website.

Tangram Vision Self-Calibrating HiFi 3D Sensor Features ROS 2 Integration and On-Device AI

Tangram Vision introduces Self-Calibrating HiFi 3D Sensor for robotics. It integrates high-fidelity 3D sensing with on-device AI for easy Depth Sensing and Edge AI.

The Team at Tangram Vision has recently introduced Self-Calibrating HiFi 3D Sensor a cutting-edge depth sensing solution designed primarily for robotics applications. The camera module integrates high-fidelity 3D sensing with on-device AI capabilities, to provide a Plug-‘n’-Play experience for Depth Sensing and Edge AI applications.

The sensor is built around the Texas Instruments Jacinto processor with a dedicated deep-learning matrix multiply accelerator (MMA) that runs at up to 8 TOPS (tera-operations per second) at 1.0GHz. The device also has 16 GB of onboard RAM with a custom-made Optiocs system.

This custom sensor system uses special optics to capture detailed depth information from its 2.2-megapixel cameras and laser projectors, covering a wide136° area. It’s incredibly precise too, with less than a one percent depth error up to 16 feet away. Plus, it automatically fixes alignment issues while it works.

Moreover, its dual connectivity options USB-C and PoEEthernet provide flexibility during development and reliability in deployment scenarios. The sensor also supports sensor fusion and offers multiple mounting options, making it adaptable to various robotics setups.

The company demonstrated the sensor’s capabilities with real-time applications like multi-target pose estimation, objectidentification, and 3Dscenesegmentation using TI’sDeepLearningLibrary(TIDL). These features, along with support for ROS2, enable easy integration into robotics projects.

HiFi 3D Sensor Features and Specifications:

  • Sensor Type: Designed for high-fidelity 3D sensing with self-calibration.
  • Integration: Seamlessly integrates with ROS 2 for robotic applications.
  • On-Device AI: Equipped with on-device AI for real-time data processing.
  • Resolution: Offers 2.2MP global shutter cameras for high-resolution imaging.
  • Field-of-View: Provides a wide 136° DFOV for comprehensive data capture.
  • Active Texturing: Utilizes dual laser pattern projectors for enhanced surface texturing.
  • Accuracy: Delivers highly accurate depth measurements with sub-1% error.
  • Neural Processing Unit (NPU): Features a powerful TI Jacinto™ processor for neural network processing.
  • Memory: Includes 16 GB of onboard memory for data storage and processing.
  • Software Applications: Compatible with TI Deep Learning Library (TIDL) for various tasks.
  • Compatibility: Native ROS 2 support for seamless integration into robotics projects.
  • Self-Calibration: Automatically recalibrates misalignments without calibration targets.
  • Sensor Fusion: Supports synchronization of multiple sensors for data fusion.
  • Dual Connectivity: Offers USB-C and PoE Ethernet options for connectivity.
  • Mounting Options: Multiple mounting options with drilled and tapped holes for easy installation.

The HiFi 3D Sensor is currently available for preorder on Kickstarter for $478.00, excluding shipping costs. For additional details, you can visit their product page.

Rockchip RK3566 Powred BIGTREETECH Pi 2 Features Gigabit Ethernet, Dual-band Wi-Fi and more

BIGTREETECH Pi 2 a RK3566 SBC with 2GB LPDDR4, 32GB eMMC, Gigabit Ethernet, dual-band Wi-Fi, Bluetooth 5.2, USB 3.0, GPIO, and more

The BIGTREETECH Pi 2 is a powerful single-board computer (SBC) powered by the Rockchip RK3566 chipset. It boasts 2GB of LPDDR4 RAM and a 32GB eMMC module, offering robust performance and ample storage capacity. With features like Gigabit Ethernet, dual-band Wi-Fi, and Bluetooth 5.2, it ensures seamless connectivity options. Additionally, the inclusion of a 40-pin GPIO header, USB 3.0, CSI, DSI, and a PCIe 2.1 interface enhances its versatility for various projects. It can be conveniently powered from a 12V24V DC input, all packed within a compact 94 x 56mm form factor.

The SBC is powered by a Rockchip RK3566 which features a Quad-core 64-bit Cortex-A55 cores running up to 1.8GHz, accompanying that is the ARM Mali G52 2EE GPU, but it lacks the NPU that we see in modern SBCs.

Previously we have covered similar SBCs like the the BeagleY-AI, the Radxa Zero 3W, the Banana Pi BPI-M2, and more feel free to check those and other SBCs out if you are looking for SBCs.

The BIGTREETECH Pi 2 Specifications

  • Rockchip RK3566 SoC: Quad-core 64-bit Cortex-A55 CPU (up to 1.8GHz) with ARM Mali G52 2EE GPU.
  • Memory/Storage:
    • 2GB LPDDR4
    • 32GB eMMC
    • 1x MicroSD card slot
  • Display/Audio:
    • 1x Micro HDMI
    • 1x 2-line DSI
    • 3.5mm Audio jack
  • Camera:
    • 1x 2-line CSI
  • Connectivity:
    • 1x 10/100/1000Mbps Ethernet
    •  Dual-band 2.4G/5G Wi-Fi
    • Bluetooth 5.2
  • Expansion:
    • M.2 Key
  • I/O Interfaces:
    • 40-pin GPIO header
  • USB:
    • 1x USB 3.0 Type-A
    • 3x USB 2.0 Type-A
  • Power:
    • 12V – 24V DC
  • Mechanical:
    • 94 x 56mm

Pricing details for the BIGTREETECH Pi 2 have not been released yet, but both products are set for official launch on April 20th. For further information and updates, visit the BIGTREETECH online store.

TOP PCB Companies