artificial intelligence News - Page 74 of 114 - CNX Software

December 9, 2020 by Abhishek Jadhav - No Comments

What is Federated Learning in 5G C-V2X?

Implementation approach of Federated learning

In the past, Bluetooth, Wi-Fi, 3G, 4G with onboard sensors and GPS gave a huge start to the partial or conditional automation of the vehicles. In today’s era of 4G/5G C-V2X (Cellular vehicle-to-everything) with onboard compute and sensors have pushed the connectivity evolution in the automotive industry. Work has increased on high-automation and fully autonomous vehicles. This is all possible due to wireless communication which is 5G. 5G brings high bandwidth, ultra-low latency, and high reliability to the board. With these promises, it becomes very evident that these technologies, specifically with 5G that can be leveraged in vehicles. The technology is not just to connect vehicles, but also tries to see if a vehicle can connect to a network to get some information from the cloud servers. For the use case of vehicles platooning, we can expect a latency of 10 ms with 99.99% reliability and a high data rate […]

December 7, 2020December 7, 2020 by Jean-Luc Aufranc (CNXSoft) - No Comments

AI Gateway combines Raspberry Pi Compute Module 4 with Google Coral Edge TPU

Techbase quickly unveiled an update to their Modberry industrial controllers with Modberry 500 CM4 DIN Rail computer as soon as Raspberry Pi Compute Module 4 was launched. The company has now introduced another Raspberry Pi CM4 based product with ModBerry AI GATEWAY 9500-CM4 gateway combining the Broadcom BCM2711C0 system-on-module to Google Edge TPU to cater to applications leveraging AI acceleration. ModBerry AI GATEWAY 9500-CM4 preliminary specifications: SoM – Raspberry Pi Compute Module 4 AI accelerator – Google Coral Edge TPU module with up to 4 TOPS connected over the RPi CM4’s PCIe interface Peripherals (customizable as needed) USB, HDMI, and Ethernet. Serial RS-232/485 ports Digital and analog I/Os with or without opto-isolation Relays, 1-Wire, CAN bus, M-Bus Master and Slave PCIe modules for wireless support Cellular – 4G/LTE and 5G modem, interchangeable with Coral TPU, plus on-board eSIM LPWAN – NB-IoT, LoRa, Sigfox, Wireless M-Bus ZigBee Additional Wi-Fi/Bluetooth interface or […]

December 3, 2020April 30, 2021 by Abhishek Jadhav - 2 Comments

Compact Jetson Xavier NX/ Nano open hardware baseboard supports Android

When it comes to NVIDIA Jetson family of modules, we should understand that NVIDIA Jetson Nano is for makers and STEM education, while Xavier NX is more geared towards professional and commercial use. The compute module Xavier NX was announced before the development kit, which includes the module and reference carrier board and otherwise for NVIDIA Jetson Nano. There are some third-party carrier boards & edge computers available for Jetson Xavier NX and Jetson Nano, including DesignCore Carrier Board and Diamond FLOYD Carrier Board. Antmicro, a custom hardware, software, and AI engineering company, came up with a compact open hardware device: Jetson Nano/ Xavier NX baseboard. It supports both NVIDIA Jetson Nano SoM as well as the Xavier NX SoM. The board is designed to put NVIDIA’s versatile module at the heart of cutting-edge projects and drive innovation in areas such as AI/ML, drones, medical and industrial object recognition systems. […]

December 1, 2020March 4, 2021 by Jean-Luc Aufranc (CNXSoft) - 46 Comments

Rockchip RK3568 processor to power edge computing and NVR applications

We recently reported on the Rockchip developer conference (RKDC) 2020, and especially the upcoming Rockchip RK3588 Cortex-A76/A55 processor that packs a lot of power and features, and is now scheduled to launch in Q3 2021. But as you can imagine there were other products revealed during the conference, including the new Rockchip RK3568, similar to the previously announced RK3566, and designed for edge computing and network video recorder (NVR) applications. There’s also an RK353X processor specifically designed for NVR application that should have similar performance as RK3568, but a lower cost, as well as two UP camera SoC with RV1106 and RV1128. CNX Software managed to obtain the presentation for the RK3568 processor from the conference so we’ll focus on the quad-core Cortex-A55 processor in this post. Rockchip RK3568 specifications: CPU – Quad-core Cortex-A55 processor with NEON and FPU GPU – Mali-G52 EE with support for OpenGL ES 1.1 to […]

November 25, 2020 by Abhishek Jadhav - No Comments

M1108 AI accelerator chip delivers up to 35 TOPS for high-end edge AI applications

Last week, Mythic announced a breakthrough with compute-in-memory technology based on a 40 nm process with what the company claims to be the industry’s first Analog Matrix Processor. The M1108 AMP AI accelerator chip targets high-end edge AI applications including smart home, AR/VR, drones, and is said to set a benchmark in the industry for high performance and low power in a single cost-effective device, also available in M.2 and PCIe form factors. The M1108 comes with an array of flash cells, ADCs, a 32-bit RISC-V nano-processor, a SIMD vector engine, SRAM, and a high-throughput Network-on-Chip (NOC) router. With 108 AMP tiles, the M1108 provides up to 35 Trillion-Operations-per-Second (TOPS) enabling ResNet-50 at up to 870 fps. This enables a power-efficient execution of complex AI models such as ResNet-50, YOLOv3, and OpenPose Body25. The industry leader NVIDIA also has a similar AI accelerator chip NVIDIA Xavier AGX which delivers up […]

November 23, 2020September 13, 2021 by Jean-Luc Aufranc (CNXSoft) - 46 Comments

ASUS unveils Tinker Board 2 SBC with faster Rockchip RK3399/OP1 processor

ASUS surprised the maker community in 2017 with the introduction of the Rochchip RK3288 powered Tinker Board to compete as Raspberry Pi 3 Model B. It was followed by Tinker Board S with built-in storage and other new features, as well as Tinker Board Edge T and Edge R SBC’s both with an AI accelerator namely Google Edge TPU and the NPU inside Rockchip RK3399Pro. The company has now launched a new model called Tinker Board 2 without AI accelerator, but featuring Rockchip RK3399, or more exactly the higher grade Rockchip OP1 used in Chromebooks, delivering 96% faster single-thread performance and a 64% boost in multi-core performance compared to the Rockchip RK3288 processor found in the original Tinker Board, while the GPU is around 28% faster with glmark2-es2 off-screen benchmark. There are two variants of the board with Tinker Board 2 and Tinker Board 2S with the latter adding onboard […]

November 23, 2020July 25, 2021 by Jean-Luc Aufranc (CNXSoft) - 13 Comments

Reolink RLC-810A review – A 4K security camera with people & vehicle detection

Last week I received Reolink RLC-810A 4K smart security camera with support for people and vehicle detection. I listed the specifications and check out the content of the package of the first security camera I’ve received with artificial intelligence. That should be great to avoid all unnecessary alerts from motion detection I get from my “dumb” IP cameras. In this review, I’ll write a small guide showing how to use the camera with the Reolink app, the web interface, and check out RTSP and ONVIF support. I’ll also see if people and vehicle detection lives up to my expectations. Reolink RLC-810A Camera installation and setup As noted in the first part of the review, the camera does not come with a power adapter and does not support WiFi. So I had to find a 12V power adapter and used a 15-meter Ethernet cable to connect it to my router. Before […]

November 21, 2020November 26, 2020 by Jean-Luc Aufranc (CNXSoft) - 9 Comments

Khadas VIM3 & VIM3L SBC’s become Android reference boards

Khadas VIM3 & VIM3L Android Reference Boards

If you need to work on the very latest development version of Android or AOSP, you can either get one of the supported phones such as Google Pixel 5, one of the reference boards for Android which should be better for development and also fairly cheaper. We previously reported about Hikey, Hikey960, and DragonBoard 845c single board computers being part of the short list of reference boards for the Android Open-Source Project (AOSP), but Amlogic powered Khadas VIM3 and VIM3L have recently been added to the official devices page which makes them the first AOSP reference boards with a dedicated NPU / AI accelerator. As a quick reminder, Khadas VIM3 is the most powerful board with an Amlogic A311D hexa-core Cortex-A73/A53 processor with up to 4GB RAM and 32GB flash, and Khadas VIM3L is the light version with an Amlogic S905D3 quad-core Cortex-A55 processor with 2GB RAM and 16GB flash. […]