artificial intelligence (AI) News - Page 39 of 121 - CNX Software

June 13, 2023June 14, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

NVIDIA Jetson AGX Orin Industrial offers ECC memory, extended temperature range, improved robustness

NVIDIA has recently introduced the Jetson AGX Orin Industrial module that comes with an extended temperature range, operating lifetime, shock and vibration specifications, as well as support for error correction code (ECC) memory, and delivers up to 248 TOPS of AI performance with power configurable between 15-75 W. The industrial-grade system-on-module is based on the Jetson AGX Orin 64GB module but with lower operating frequencies to enable a longer operating lifetime and improved endurance. It’s form-factor and pin-compatible with Jetson AGX Orin, and NVIDIA says it delivers more than eight times the AI performance of the 30 TOPS Jetson AGX Xavier Industrial introduced in 2021, but does not integrate Cortex-R5F cores. NVIDIA Jetson AGX Orin Industrial specifications: CPU – 12-core Arm Cortex-A78AE v8.2 64-bit processor @ 2.0 GHz with 2MB L2 + 4MB L3 cache GPU / AI accelerators NVIDIA Ampere architecture with 2,048 NVIDIA CUDA cores and 64 Tensor […]

June 9, 2023July 28, 2023 by Jean-Luc Aufranc (CNXSoft) - 5 Comments

A first look at i-Pi SMARC 1200 (Genio 1200) development kit

ADLINK Technology’s i-PI SMARC 1200 is an industrial prototyping kit for AI IoT and Robotics based on a SMARC 2.1 compliant LEC-MTK-I12000 system-on-module (SoM) powered by the MediaTek Genio 1200 (MT8395) octa-core Cortex-A78/A55 AIoT processor with a 4.8 TOPS AI processor. The company sent me a sample for evaluation and review, and in this first post, I’ll go through an unboxing and how to boot the system with the supported Yocto Linux or Ubuntu. i-Pi SMARC 1200 unboxing The i-Pi SMARC 1200 may be destined for industrial users, but it still comes with a neat retail package. The “i-Pi LEC-MTK-I1200-44-64G/EU” kit features a CPU module with 4GB LPDDR4 memory and 64GB UFS storage. The package includes the board itself preassembled in a desktop frame, a Quick Start Guide, a 19V/2.37A (45W) power adapter with an EU power cord, a micro USB to USB cable for debugging, and two bits for […]

June 9, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

Axiomtek AIE110-XNX is an affordable Edge AI Developer Kit based on NVIDIA Jetson Xavier NX

Axiomtek AIE110-XNX Edge AI developer kit provides up to 21 TOPS of accelerated AI computing performance thanks to an NVIDIA Jetson Xavier NX system-on-module with a 6-core NVIDIA Carmel ARM v8.2 64-bit CPU and a 384-core NVIDIA Volta GPU with 48 Tensor Cores. The new devkit serves as a relatively low-cost cost development kit with no minimum order, instead of getting full-feature embedded computers such as the company’s AIE900-XNX Edge AI system also based on the same Jetson Xavier NX module. Axiomtek AIE110-XNX specifications: NVIDIA Jetson Xavier NX system-on-module with CPU – 6-core NVIDIA Carmel Armv8.2 64-bit CPU with 6 MB L2 + 4 MB L3 cache GPU – 384-core NVIDIA Volta GPU with 48 Tensor Cores AI Accelerator – 2x NVDLA System Memory – 8GB 128-bit LPDDR4x onboard Storage – 16GB eMMC flash Storage – M.2 Key M 2280 with PCIe x4 NVMe SSD slot, microSD slot Video Output […]

June 2, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

Khadas VIM4 Amlogic A311D2 SBC gets 3.2 TOPS NPU

First introduced in 2022, the Khadas VIM4 SBC will now sell with an Amlogic A311D2-N0D Cortex-A73/A53 processor equipped with a 3.2 TOPS NPU, and the version without AI accelerator will be phased out. I reviewed the Khadas VIM4 with both Android 11 and Ubuntu 22.04 last year, and really liked the user experience notably thanks to the OOWOW system that makes it really easy to try different operating systems downloaded from Khadas servers, although some features like support for GPU acceleration still needed to be worked on at the time. One thing missing compared to the previous generation Khadas VIM3 SBC was an on-chip NPU to accelerate AI workloads, and the latest revision of the Amlogic A311D2 processor (with N0D suffix) fixes that on the Khadas VIM4. “Khadas VIM4 with NPU” specifications: SoC – Amlogic A311D2-N0D octa-core processor with 4x Arm Cortex-A73 cores @ up to 2.2 GHz and 4x […]

May 25, 2023May 25, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

Reolink E1 Outdoor Pro 4K smart security camera supports WiFi 6 connectivity

Reolink E1 Outdoor Pro is a 4K smart security camera with PTZ and auto-tracking like the Reolink TrackMix, but instead of relying on PoE or WiFi 4/5, the camera comes with WiFi 6 connectivity for a faster and more stable network connection with less latency. Most of the WiFi security cameras I’ve been using come with 2.4GHz WiFi 4, and I’ve found connectivity to the solar-powered Reolink security cameras I’ve been using to be somewhat unreliable with difficulty connecting to them through the mobile app, although uploads to the cloud don’t suffer from those issues. So I typically prefer PoE or 4G LTE security cameras since the connectivity is more stable, but Reolink E1 Outdoor Pro security camera’s WiFi 6 support could help depending on the environment. Reolink E1 Outdoor Pro specifications: Camera – 4K Ultra HD resolution with 3x optimal zoom Color night vision with spotlight PTZ – 355º […]

May 25, 2023May 24, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

iW-RainboW-G55M is an OSM-LF compliant module based on TI AM62A Cortex-A53 processor

iWave Systems iW-RainboW-G55M is an OSM Size L compliant system-on-module based on Texas Instruments AM62A single to quad-core Cortex-A53 processor with up to 8GB RAM, 128GB eMMC flash, and a wireless module with WiFi 6, Bluetooth 5.2, and 802.15.4 radios. The iW-RainboW-G55M can leverage the vision processing and deep learning accelerator and the Arm Cortex-R5F real-time cores for control and device management found in the AM62A processor, as well as its display and camera interfaces, peripheral and networking options, to develop products for “man-machine applications at the edge”. iW-RainboW-G55M SoM specifications: SoC – Texas Instruments AM62A SoC (AM62A3, AM62A3-Q1, AM62A7, or AM62A7-Q1) Application processor – Up to four Arm Cortex- A53 cores @ 1.4GHz Real-time cores 1x Cortex-R5F @ 800MHz (MPU Channel with FFI) 1x Cortex-R5F @ 800MHz to support Device Management C7xV-256 Deep Learning Accelerator up to 2 TOPS System Memory – 2GB (default) to 8GB LPDDR4 RAM Storage […]

May 19, 2023May 18, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

Cincoze DX-1200 fanless Alder Lake-S embedded computer targets machine vision, industrial automation

Cincoze DX-1200 is a fanless embedded computer powered by a choice of 12th gen Intel Core (Alder Lake-S) Celeron up to Core i9 CPUs whose mainboard is housed in a rugged and compact chassis designed for industrial edge AI and machine vision applications. The DX-1200 supports up to 64GB of 4800MHz DDR5 memory with ECC, and the Intel Xe architecture of the UHD 770 graphics chip is said to deliver image classification inference performance that is 2.8 times the speed of the Comet Lake-S platform. The embedded computer also offers a wide range of I/Os and as well expansion capabilities for up to four 10GbE, DIO, and COM ports, and M.2 and mPCIe sockets for WiFi, GNSS, 4G, and Bluetooth connectivity. Cincoze DX-1200 specifications: Alder Lake-S SoC (one or the other) Intel Core i9-12900E 16 cores up to 5 GHz with Intel UHD Graphics 770; TDP: 65W Intel Core i7-12700E […]

May 18, 2023May 18, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

AndesAIRE AnDLA I350 AI/ML IP block is configurable from 64 GOPS to 8TOPS for Edge AI SoCs

Andes Technology has just announced the AndesAIRE product line, where AndesAIRE stands for Andes AI Runs Everywhere, comprised of the AndesAIRE AnDLA I350 (Andes Deep Learning Accelerator) AI/ML hardware accelerator intellectual property (IP) and the AndesAIRE NN SDK with neural network software tools and runtimes. AndesAIRE AnDLA I350 AnDLA I350 specifications: Configurable MACs from 32 to 4096 (INT8) Maximum performance – 8 TOPS at 1GHz Configurable local memory – 16KB to 4MB Multi-dimension DMA Four 64-bit AXI bus interfaces NN type – CNN inference NN models Image and Video: AlexNet, VGG-16/19, MobileNet-v1/v2/v3, ResNet-8/50, Tiny YOLO v1/v2, YOLO v1/v2/v3/v4/v5, SSD MobileNet v1/v2, Inception v2, EfficientNet-lite, MobileFaceNet, BlazeNet Speech/Voice and audio: LSTM, RNN, GRU Operators: Conv2d, depthwise convolution, pointwise convolution, transpose convolution, dilated convolution, element-wise (add, sub, mul), fully-connected, activation (ReLU, leaky ReLU, sigmoid, Tanh, ReLU6, SiLU), pooling (max, ave), upsample, concatenation, batch normalization, channel padding Operator fusion NHWC data format The IP […]