Android APP Contact Us
login / register Upload
TRENDING:

Axelera AI M.2 is the game-changing Accelerator Module in the Field of Machine Learning

Posted on by Saumitra Jagdale

As the demand for AI-based deep learning applications increases there is a demand for AI accelerators with higher processing capacity. An AI accelerator is a high-performance parallel computation machine that is specifically designed to perform tasks with the help of neural networks. One such revolutionary accelerator is Axelera AI M.2. Fueled by a single Metis AIPU, this module can deliver up to 214 tera-operations per second (TOPs) while consuming very minimal power as low as 15 TOPs/W. Four homogenous AI cores are equipped in the AIPU. The cores are designed for accelerating neural network inference.

Technical Specifications of Axelera AI M.2

  • Axelera AI M.2 is based upon the company’s Metis AIPU(AI Processing Unit) which is designed in 12nm CMOS.
  • The AI core, a RISC-V controlled dataflow engine, delivers an unprecedented peak performance of 214 TOPs in total. The module has an energy efficiency of 15 TOPs/W at INT8.
  • INT8 represents an 8-bit signed integer which holds values from -128 to 127—typically used in deep learning applications due to its fast matrix operation.
  • The 214 TOPS with 15 TOPs/W gives the Axelera AI M.2 a great advantage against its rivals like Myriad 2, Hailo-8 and NVIDIA Jetson AGX Orin module.

Extending processing capabilities with versatile connectivity

The device connects to the accelerator module using an M.2 slot, which replaces mSATA slots. M.2 offers a more versatile and suitable interface for solid-state storage, especially in smaller devices. M.2 provides a range of module widths and lengths, and advanced interfacing features, whereas mSATA is limited in comparison.

The Axelera Metis is programmable by the Voyager SDK which provides an end-to-end integrated framework for Edge applications development. Developers optimize the SDK for creating computer vision applications for the Edge because it is specifically designed for AI-related tasks.

The company is also launching a PCIe card for developers seeking more performance. It will feature four Metis AIPUs which will be capable of providing 856 TOPS. These AIPUs can process ResNET50 at 12,800 fps and MobileNet V2-1.0 at 38,884 fps. This provides a substantial boost in the field of neural networks. The company additionally teased their AI gateway with Metis M.2 card. It is a solution for businesses looking to deploy networks of smart cameras. AI vision gateway offers performance in excess of 100 TOPs for small to medium-sized smart camera deployments in robotics and retail (up to 8 cameras).

The M.2 accelerator module retails at $149 and the PCIe card costs $499. The cost of the gateway is not yet available. For more information visit the product site.

Walkie-Talkie module Built On Tiny ESP32 board

Posted on by Rakesh Kumar, Ph.D.

Recently, LILYGO introduced a compact ESP32-S3 gadget that features a Walkie-Talkie module along with additional components like an OLED screen, speaker, microphone, and a slot for a 18650 battery. An ESP32 gadget with a stylus-equipped LCD display was also on show.

The T-TWR is said to have an embedded ESP32-S3-WROM-1-N16R8 microprocessor, as detailed on the product page. Xtensa’s 32-bit LX7 dual-core ESP32-S3 processor (up to 240MHz) and 16 MB of flash memory. With Bluetooth 5.0 and Wi-Fi at 2.4GHz, the PSRAM has 8MB of storage.

T-TWR pinout

A 0.96-inch OLED LCD display that supports the I2C interface protocol is built within the system; this model is the SSD1306. The built-in walkie-talkie module supports 1.6w/1.8w  and operates at a bandwidth of 12.5KHz/25KHz. The module can also be purchased in additional radio-frequency options, such as UFV, VHF, and 350. In addition to one suitable antenna and two 13-pin headers, the order appears to contain a T-TWR ESP32-S3 module. The necessary information can be found in the GitHub repository linked below.

T-TWR module

T-TWR features a 0.96-inch OLED SSD1306 screen (128 x 64). With a sensitivity of -124 dBm and a bandwidth of 12.5 kHz/25 kHz, the walkie-talkie module is compatible with frequencies in the ranges of 400 – 480 MHz UFV (optional), 134 – 174 MHz VHF (optional), 320 – 400 MHz, and 350 MHz (optional). There is one USB Type-C port available. There is only one microphone and one speaker for input and output, respectively. The status LEDs and IO3, and stemma QT/ Qwicc connectors, as well as the encoder, are all included. This item has a size of 122 x 32 x 22 mm. Battery operation is used.

T-HMI pinout

T-HMI embedded board is also based on ESP32-S3R8 microprocessor. Xtensa 32-bit LX7 dual-core (up to 240MHz) with 16 MB Flash and 8 MB PSRAM is featured in the ESP32-S3. Bluetooth 5.0 and Wi-Fi at 2.4 GHz are also included. The screen is a 2.8″ resistive IPS TFT LCD (ST7789V) with a resolution of 240 pixels by 320 pixels. Similar to other TFT LCDs, this one communicates via a parallel 8-bit interface. The board has a USB Type-C port for external power and programming, as well as a battery connector for use with an external power source, as can be seen in the images below.

T-HMI peripherals

On the product page, there is a GitHub repository with examples that can be used as documentation. For this product, the order comes with the board, a resistive screen stylus pen, and a PH2.0 battery cable. The T-TWR board can be bought on AliExpress.com for $45.58. In the same way, LILIGO’s online store lists the T-HMI for $23.73.

Highlights of RSC201 Fanless Edge Computer with Xilinx Kria SoM

Posted on by Rakesh Kumar, Ph.D.

Based on Xilinx’s Kria K26 System-on-Module, Axiomtek’s RSC201 Vision System was released last month for use in business and industry. With a single 1x GbE LAN port, a single 1x CAN bus, a single 1x 4K display, and 5G connectivity, the RSC201 is well-equipped. An H.264/265 video codec is built right into the Kria K26 SoM, which also boasts a Quad-core Arm Cortex-A53 MPCore (up to 1.5GHz), a Dual-core Arm Cortex-R5F MPCore (up to 600MHz), a Mali-400 MP2 (up to 667MHz), and 1.36TOPS.

Block diagramKria K26 SoM

The device has one M.2 Key B 2242 slot and up to 16GB of eMMC flash storage for SATA 3.0. There is only one DisplayPort for this screen, and its maximum resolution is 4Kp60. The RSC201 has an M.2 B-Key 3052 slot for 5G modules, a SIM slot, and a Gigabit Ethernet interface. The RS201 is approved by CE and FCC as Class A and is immune to vibration (3.5 Grms, 5 to 500Hz, X, Y, Z directions). Although the product page boasts Linux compatibility, as of this writing, the Downloads page is still empty.

RSC201 peripherals

 

About 64 bits of onboard DDR4 (Up to 4GB) and eMMC (Up to 16GB) are included in the RSC201. There includes a GbE LAN port for communication, and the DisplayPort version is 1.4. There is a single SIM card slot and one each M.2 Key B 2242 and M.2 B-Key 3052 expansion slots. There are two USB 3.0 ports. There is one RS-232 COM port, one PMOD port, and one CAN bus port used for input and output. Linux is the operating system, and there are 2 LEDs for monitoring system activity.

The package weighs roughly 1.5 kg and has dimensions of 150.0 x 111.0 x 65.6 mm. Aluminum extrusion and the robust steel construction also contribute to its IP40 rating. You may use it from -30 degrees to 75 degrees, and the voltage is between 12 and 24 volts.

Cameras using the GbE LAN interface are supported, and so is the 5G wireless connection. Non-condensing relative humidity between 5% and 95%. The X, Y, and Z axes vibrate at frequencies between 5 and 500 Hz, and the overall level of vibration is about 3.5 Grms. The option to place it on the wall is included. Its cutting-edge capabilities make it a top choice for use in edge AI smart city and industrial settings.

Refer product page for further information.

M0S Module TinyML Supported By Affordable RISC-V BL616

Posted on by Rakesh Kumar, Ph.D.

Sipeed released two embedded devices based on Bouffalo Lab‘s RISC-V BL616 microcontroller. The M0S module has interfaces for WiFi 6, Bluetooth 5.2, and Zigbee. It also works with DVP cameras, RGB LCDs, and Ethernet RMII.

On the product page, it says that the BL616(RV32GCP) has a BL6(RV32GCP) chip with a 32-bit RISC-V CPU with FPU and DSP unit (up to 384MHz) and 4 MB Flash. It has 480KB SRAM and L1 Cache as well.

BL616 block diagram

BL616 system architecture 

According to the datasheet, the MCU also has support for JTAG, XIP, and QSPI flash, in addition to two 32-bit general-purpose timers and an RTC timer with a maximum count of one year.

M0S module & pinout

The MCU can sample at rates of 8, 12, 16, 22,05, 24, 32, 44,1, and 48 kHz, respectively. The BL616 has the following capabilities for security: secure booting, secure debugging, XIP QSPI on-the-fly AES decryption, support for TrustZone, support for AES-CBC/CCM/GCM/XTS, a TRNG, and a Public Key Accelerator for RSA/ECC. All of these capabilities are implemented in hardware.

The product page features some available resources, such as the Bouffalo Lab Software Development Kit (SDK), the Sipeed Wiki, which appears to be empty as of the publication date, and a GitHub repository. The datasheet for the BL616 may be obtained in the GitHub repository maintained by Bouffalo Lab and is also accessible through the repository maintained by Sipeed.

M0S Dock

The M0S module has a wide variety of features, including an RGB LCD and a DVP camera display. Connectivity options include Ethernet RMII, 2.4GHz Wi-Fi 802.11b/n/ax., BT/Bluetooth of 5.2, 802.15.4 (Zigbee/Thread), and an audio ADC (MIC, SNR>92dB) and audio DAC (Speaker, SNR>95dB). It supports USB on the go (OTG) from USB 2.0 devices (up to 480Mbps). TinyMaix, MaixHub, and the I/O protocols I2C, UART, and SPI are all included in the software. 10mm x 11mm is the size. Ultra-low power mode (1uA) at 3.3V provides power.

You can purchase a Sipeed M0S module for $3.19 On AliExpress, while the Dock model costs $4.39. There is also a $1.20 off one-item coupon that may be used at the store.

Featuring Arduino Nicla Voice with Speech Recognition, BL5.0 and Much More

Posted on by Rakesh Kumar, Ph.D.

Recently, Arduino has introduced a small, low-power integrated module with AI capabilities. Syntiant NDP120 processor designed for Deep Learning applications and ANNA-B112 u-Blox module for wireless communication forms the basis of the Arduino Nicla Voice. The Nicla Voice, much like the recently highlighted Nano 33 BLE Sense Rev2, features an ANNA-B112 module built on the Nordic nRF52 SoC series and a Syntiant Neural Decision Processor.

One Syntiant Core 2 U-LP deep neural network inference engine, one Tensilica HiFi 3 Audio DSP, and one Arm Cortex M0 (up to 48MHz), 48KB of static random access memory (SRAM) are all included in the Syntiant NDP120.

The nRF52832 features an Arm Cortex M4 processor (up to 64 MHz), 64KB of SRAM, and 512KB of Flash-mapped FIFOs. Further, it has a 12-bit/200 ksps ADC, 2.4-2.4835 GHz BL 5.0 (through the ARM Cordio stack), and 4.2 (via ArduinoBLE). Not only that but there’s also a 32MHz oscillator on the inside and a 1.8V power supply.

 

NDP120 block diagram

ANNA B112 block diagram

This datasheet clearly states that  NDP120 supports “concurrent neural networks, 2D convolution, depth-wise convolution, the recurrent neural network including LSTM and GRU, average and max pooling”.

According to Hackster.io, the Arduino Nicla Voice will cost about $82. To sign up for the waiting list, look at the product announcement. On the Hardware page for the Arduino Nicla Voice, you can find the datasheet, libraries, schematics, STEP files, and other guides.

Arduino Nicla Voice pinout

It appears that the Nicla Voice, like the Nano 33 BLE Sense Rev2, uses IMUs manufactured by BOSCH. These IMUs consist of a 6-axis gyroscope (BMI270) and a 3-axis accelerometer (BMM150).

Arduino talks about its ability to recognize speech,

“the Nicla Voice speeds up the tempo of product development, whether your launch requires solutions that use voice to play music, make phone calls or pull up the blinds. It makes devices smarter by allowing for multiple wake words, replacing buttons users struggle with, and helping designers integrate unobtrusive voice-command systems into stylish accessories”.

The Arduino Nicla Voice has a list of specs, such as 16 MB SPI Flash storage, BLE 5.0, and Bluetooth mesh connectivity. I/O Interfaces are done through GPIOs 1.8V/3.3V, 10x Digital I/Os, 12x PWM pins, 1x I2C bus (with ESLOV connector), 1x UART, 1x SPI, and 2x ADC. The 6-Axis IMU (BMI270) and 3-Axis IMU (BMM150), 1x Digital MEMS Microphone (IM69D130), and 1x RGB LED are some of the other features. The USB is a Micro-B port. The temperature range is 0°C to 70°C. It gets its power from a 3.3V USB-B Micro port, a 3.7V Li-Po battery, and a JST connector. The size and weight are both 22.86 x 22.86 mm and 2g.

High Geared RISC-V ESP32-P4 SoC Loaded with GPIOs by Espressif

Posted on by Rakesh Kumar, Ph.D.

The ESP32 family of system-on-chips (SoCs) from Espressif now includes a new member. The RISC V-based ESP32-P4 has three processing cores, fifty general-purpose input/output (GPIO) pins, and a full complement of security options. Interestingly, though, the chip lacks the RF radio.

Despite having three CPU cores, Espressif set them up as a “big-little” architecture with a high-performance (dual-core) and low-power (single-core) setup, respectively. When they aren’t needed, high-performance cores should be turned off.

The ESP32-P4‘s high-performance central processing unit (CPU) is a 400 MHz dual-core RISC-V CPU. The SRAM on the chip is 768 kilobytes in size. When paired with an external PSRAM, however, the on-chip RAM acts as a cache for the immediate area. For low latency buffer access, there are also eight kilobytes of zero-wait tightly linked memory (TCM).

One RISC-V core with up to 40 MHz of clock speed plus specialized SRAM, ROM, and peripherals make up the low-power system. It’s also where the system-on-power chip’s management circuitry is located (PMU). It has a touch interface, a temperature sensor, and low-speed serial interfaces as its peripherals.

The ESP32-P4’s superior security is one of Espressif’s main selling points. Secure boot, flash encryption, a genuinely random number generator, digital signature management, access rights, and privileged isolation are all part of the package.

The 50 general-purpose I/O pins are the most ever included in an Espressif system-on-chip. MIPI (Camera and Display) ports, hardware accelerators for h264 and JPEG, USB, Ethernet, and many more are among the various peripherals available for high-performance CPUs. User interface options on the ESP32-P4 include capacitive touch, speech recognition, and CSI and DSI interfaces. One can use a Pixel Processing Accelerator, which is a piece of hardware, to speed up the processing of graphical displays.

The ESP32-P4 is the first member of the ESP32 series without a radio frequency (RF) transceiver of any type. Wi-Fi, Bluetooth, and other 2.4 GHz wireless technologies are not included. Instead, the ESP32-P4 can communicate with other devices in the ESP32 family using ESP-Hosted, ESP-AT, or a third-party solution, as claimed by Espressif. Wire-based Ethernet is also an option. Overall, the ESP32-P4 is made for edge computing applications that require powerful processing, a sophisticated human-machine interface, and robust security.

The Espressif-IDF will, of course, support the brand-new ESP32-P4 SoC. They promised additional ESP32-P4 details during the next three weeks, and they delivered. Please see the Espressif statement for further information.

Implementing the 100 MHz 6502 In an FPGA

Posted on by Tope Oluyemi

Jürgen Mülle has posted details of re-implementing the 65C02 in an FPGA, in a pin-compatible format that enables you to upgrade old computers and games to a 100 MHz clock rate! Speaking about the project, Jürgen says:

“The idea of implementing a CPU core inside an FPGA is not new, of course. In fact, the CPU core I am using is not my own, but was developed as a 6502 core by Arlet Ottens, and extended to cover the 65C02 opcodes by Ed Spittles and David Banks.”

He started by designing a circuit board that is the same size as a 40-pin dual inline package and has pins that correspond to the 65C02 pinout. This board was then fitted with a Spartan-6 FPGA and 64 kByte of on-chip RAM. He implemented logic within the FPGA that is able to interact with the external 65C02 bus in accordance with the clock signal from the host system. The CPU core within the FPGA clocks at a speed of 100 MHz. The goal is to create a single accelerator that can be used with 6502 and 65C02-based host computers, simply by plugging it into the CPU socket.

The FPGA board must be aware of the memory map of its host in order to determine where the host has memory-mapped I/O. It can store up to 16 different memory maps and switch between them using a mini DIP switch. When the 65F02 is turned on, it takes all of the RAM and ROM content from the host and stores it in its own internal RAM, excluding the I/O area. Once the CPU is running, the internal memory is accessed at 100 MHz for all bus operations, except for I/O addresses. In this case, the CPU will pause and an external bus cycle will be initiated at the external clock rate.

Supported host systems


One of the host systems the project supports are the Chess computers. Jürgen Mülle primarily concentrated on Mephisto chess computers, because they are the main focus of Roland’s collector’s desires. However, he hopes to extend the project to many other items on the wishlist. including Fidelity (Elite A/S Budapest and Glasgow, Prestige, Avantgarde), Novag (Super Constellation, Super Expert C), SciSys (Turbostar),

For microcomputers, the project supports the two 6502-based members of the classic 1977 “Trinity” of home computers – the Apple II and the Commodore PET series. Also, in the future, it will be possible and worthwhile to offer support for some slightly later machines like; The Acorn BBC Micro, Atari 400 and 800, and possibly the Commodore C64 come.


Note that this is a hobby project, and Jürgen Mülle has no plans at the moment to take it commercial. Some of the small-form-factor PCBs have been assembled, and have been successfully tested in the above-mentioned computers. For more information, visit the project page.

Adafruit Feather AIoT Board with ESP32-S-3,WiFi, BLE, and CircuitPython Compatibility

Posted on by Saumitra Jagdale

The Internet of Things (IoT) is a network of physical devices, vehicles and other items embedded with electronics, software, sensors, and connectivity which enables these objects to connect and exchange data. The Adafruit Feather AIoT board is a powerful and versatile tool for building Internet of Things (IoT) projects. Equipped with an ESP32-S3 microcontroller, this board offers a combination of WiFi, BLE, and low-power capabilities, making it well-suited for a wide range of IoT applications.

Technical Specifications

  • The Adafruit Feather AIoT S3 is powered by Espressif’s ESP32-S3 WROOM 1 N8R8 module. The ESP32-S3 uses the Xtensa dual-core LX7 processor which is based on 32-bit architecture and works at 240MHz.
  • The board features 384 KB ROM, 8MB Flash, 8MB PSRAM and 512KB SRAM and comes up with Wi-Fi 4 and Bluetooth 5.
  • You can charge LiPo and Li-Ion batteries on the Adafruit Feather AIoT S3 board using the 5V USB Type-C port. The board supports I2C, SPI and UART interfaces for connecting to external devices.

The ESP32-S3 microcontroller’s vector extension makes it suitable for machine-learning applications like image and signal processing, object detection and recognition, and neural network interference.

Connectivity Features of Adafruit Feather

The Adafruit Feather AIoT S3 features Wi-Fi 4 and Bluetooth 5 for wireless communication. A 2-pin connector is available on board for a single Cell LiPo / Li-Ion battery. 3 buttons are available on board for reset, boot, and user buttons. Pressing the reset button once will restart the ESP32-S3 processor while pressing it twice will enter UF2 Bootloader mode. Additionally BOOT button, when pressed, grants access to the ROM Bootloader, which can be utilized with esptool.

The Maker Feather AIoT S3 development board supports several programming languages and environments, including Arduino, CircuitPython, MicroPython, and Espressif IDF. This makes it a versatile and flexible option for a wide range of Internet of Things (IoT) projects.

Some applications include smart home automation, environmental monitoring, voice-controlled smart devices, location tracker, and many more. Furthermore, with the Adafruit Feather AIoT S3, you can control and monitor various devices and appliances in a smart home, such as lights, thermostats, and security cameras. With the Adafruit Feather AIoT S3, you can collect temperature, humidity, air quality and several other factors in the environmental monitoring field.

You can purchase the Adafruit Feather AIoT S3 for $19.95 without a header or for $21.20 with a header. For more information visit the product page. To buy the product visit Cytron Marketplace.

The UP Xtreme i12 Edge Accelerates AI Application Deployment with a Compact, Fast & Powerful Edge System

Posted on by mixos

The UP Xtreme series meets 12th Generation Intel® Core™/Celeron® power with the release of the UP Xtreme i12 Edge.

AAEON, a leading manufacturer of AI development solutions, announces the UP Xtreme i12 Edge, its latest addition to the UP product series.

The UP Xtreme i12 Edge makes substantial upgrades on its predecessor with the power-efficient multicore architecture of 12th Generation Intel® Core™ i7/i5/i3/Celeron® Processors (formerly Alder Lake-P), onboard LPDDR5, and AI module support all contributing to its potential as a system suited to industrial, healthcare, and transport applications.

Providing exceptional expansion options, the UP Xtreme i12 Edge has four M.2 Keys for Wi-Fi 6, 5G, AI, and PCIe 4.0 storage module support. With such options, users are able to bring edge AI computing capabilities to a wider variety of markets. Further, the UP Xtreme i12 Edge supports Intel® Iris® Xe graphics, Intel® DL Boost, and the Intel® Distribution of OpenVINO™ Toolkit, boosting GPU image classification inferencing performance by up to 2.81 times that of the previous generation.

Making use of its fanless 152mm x 123.8mm x 72.5mm chassis, the UP Xtreme i12 Edge is equipped with one USB 4.0, three USB 3.2, and three USB 2.0 ports, along with an Intel® i226-IT LAN port running at 2.5Gpbs with Time-Sensitive Networking (TSN). With such a combination, AAEON claims the system can achieve real-time data processing thanks to its onboard LPDDR5 system memory.

Its numerous I/O ports compliment the UP Xtreme i12 Edge’s display interface, which can support four simultaneous 4K displays via configurations of HDMI 2.0, DP 1.4a, and eDP 1.4b ports, further enhanced by Intel® Iris® Xe graphics. These features lead AAEON to believe the UP Xtreme i12 Edge will be a popular choice for smart retail and industrial automation solutions, particularly given its fanless chassis and 12~36V power input range.

The UP Xtreme i12 Edge is now in mass production and available for order, with prices dependent on SKU.

For more information about the UP Xtreme i12 Edge, please visit our product page or contact an AAEON representative directly.

LinkStar-H68K – The Mini Computer Positioned as a Router

Posted on by Emmanuel Odunlade

Introduction

IoT hardware specialist, Seeed Studio, recently released a very impressive ‘Router’ (more than a router actually) and we’ll like to take an intensive look at it. So for this article, we will be reviewing the LinkStar-H68K multimedia router and discussing the features that make it stand out amongst others.

The first thing I noticed about the ‘router’ is its compact size. It is light and small; something you can actually lift with one hand and carry everywhere. It is also enclosed in a portable metal box for better heat dissipation.

The LinkStar-H68K Router is built with convenience and flexibility in mind. It comes with enough expansion ability and media player functionality, presenting you with a fully functional device that you can be used at home, in the office or just anywhere you travel.

At the heart of the router is Rockchip RK3568 quad-core Cortex-A55 processor with an ARM Mali-G52 2EE graphics and a Neural processing unit for up to 0.8 TOPS AI performance. The device is also equipped with up to 32GB of eMMC flash storage along with one SD card socket and one USB 3.0 Type-C port that supports an external hard drive for massive storage expansion. There’s also a USB 3.0 Type-A and two USB 2.0 Type-A ports for further usage.

The device can also be used as a neat network player to instantly stream and display media given that it outputs up to 4K immersive visuals. You can also use it to access online services like YouTube or Spotify.

Overall, the LinkStar-H68K Router has up to 4x Ethernet interfaces for 2x 2.5G and 2x 1Gb, so you can use it to establish network-attached storage to store and retrieve abundant information over a network with physical file transformation. The router doubles also as a rugged mini-computer that could be used for digital signage and other applications.

Another impressive feature of the LinkStar-H68K router is the outstanding dual-band Wifi 6 which offers faster throughput speeds and less bandwidth congestion. It can perform as a repeater connecting different WiFi at different places and linking all your devices, which is beneficial for your travel or business trip.

Specifications

  • Rockchip RK3568 quad-core Cortex-A55 processor @ up to 2.0 GHz
  • Arm Mali-G52 2EE GPU
  • 0.8 TOPS AI accelerator
  • 4Kp60 H.265/H.264/VP9 video decoder
  • 1080p60 H.264/H.265 video encoder
  • 4GB of LPDDR4X memory
  • 32GB of eMMC flash storage
  • 1x microSD card socket
  • 1x HDMI 2.1 port up to 4Kp60
  • 3.5mm audio jack
  • 2x 2.5Gb Ethernet RJ45 ports via 2x Realtek RTL8125B tested up to 2.35/2.31 Gbps
  • 2x 1Gb Ethernet RJ45 ports via 2x Realtek RTL8211F tested up to 944/941 Mbps
  • Optional Wifi 6 up to 1200 Mbps
  • Bluetooth 5.2 via MT7921 PCIe module and two external antennas
  • 1x USB 3.0 Type-C port
  • 1x USB 3.0 Type-A port
  • 2x USB 2.0 Type-A ports
  • Power button and IR receiver
  • 3x LEDs
  • Pin holes for firmware updates, factory reset and reboot
  • Power Supply: 5 to 24V DC via power adaptor or 5V via USB Type-C connector
  • Power consumption: 8W
  • Temperature range: -10 to 55 degrees celsius
  • Dimensions: 80 mm x 60 mm x 40 mm

Highlights of the LinkStar-H68K Multimedia Router

  • Small and compact; it can be lifted with one hand and carried everywhere
  • Great wireless capability and high throughput performance with Bluetooth 5.2 and an outstanding WiFi 6 technology
  • Fast-speed router deployment
  • High storage capacity and great extensibility
  • 4K multimedia player
  • Support for multiple operating systems including Ubuntu, OpenWRT, Android 11, etc.

The router comes pre-installed with Android 11 and has support for Ubuntu, Debian, OpenWrt, Android, Armbian, and Buildroot. The router is offered in two versions; an entry-level model without a wireless card and a more expensive version with MediaTek MT921 wireless module. Since the Android and Ubuntu images do not support the wireless module, you may have to switch to OpenWrt or Debian or just use your own wireless solution.

You can charge it via a 5V to 24V power barrel Jack or or a 5V via USB Type-C port. The router can also be operated between -10 to 55 degrees Celsius.

Conclusion

Seeed Studio is offering the LinkStar-H68K Router for $129 but you can buy it for less at $119 if you pre-order before the end of this month (December 30th). The price mentioned above is actually for the version with 4GB RAM/32GB eMMC storage with WiFi 6 but the one with 2GB RAM/32GB eMMC and without WiFi 6 sells for $99 (same $10 early bird discount applies).

TOP PCB Companies