artificial intelligence News - Page 25 of 100 - CNX Software

January 10, 2023January 11, 2023 by Jean-Luc Aufranc (CNXSoft) - 5 Comments

Maker Feather AIoT S3 – An ESP32-S3 board programmable with CircuitPython or Arduino

Cytron Maker Feather AIoT S3 is an ESP32-S3 board compatible with the Adafruit Feather form factor and suitable for makers and STEM education with features like LEDs for GPIOs, a buzzer, expansion headers and connectors, and support for CircuitPython & Arduino. The WiFi and Bluetooth LE IoT board also supports LiPo and Li-Ion batteries, includes a USB Type-C port for power/charging and programming, and a few buttons. It’s suitable for machine learning thanks to the vector extensions found in the ESP32-S3 microcontroller. Maker Feather AIoT S3 specifications: Wireless module – Espressif ESP32-S3-WROOM-1-N8R8 module with ESP32-S3 dual-core LX7 microprocessor @ up to 240 MHz with Vector extension for machine learning, 8MB FLASH, 8MB PSRAM, WiFi 4 and Bluetooth 5 LE/Mesh USB – 1x USB Type-C port power and programming Expansion 2x GPIO headers compatible with Adafruit Feather form factor and FeatherWing add-on boards 3x Maker Ports compatible with Qwiic, STEMMA QT, […]

January 10, 2023March 5, 2023 by Jean-Luc Aufranc (CNXSoft) - 2 Comments

reComputer J4012 mini PC features NVIDIA Jetson Orin NX for AI Edge applications

reComputer J4012 is a mini PC or “Edge AI computer” based on the new NVIDIA Jetson Orin NX, a cost-down version of the Jetson AGX Orin, delivering up to 100 TOPS modern AI performance. The mini PC is based on the Jetson Orin NX 16GB, comes with a 128GB M.2 SSD preloaded with the NVIDIA JetPack SDK and offers Gigabit Ethernet, four USB 3.2 ports, and HDMI 2.1 output. Wireless connectivity could be added through the system’s M.2 Key E socket. reComputer J4012 / J401 specifications: SoM – NVIDIA Jetson Orin NX 16GB with CPU – 8x Arm Cortex-A78AE core @ up to 2.0 GHz with 2MB L2 + 4MB L3 cache GPU/AI 1024-core NVIDIA Ampere GPU with 32 Tensor Cores @ up to 918 MHz 2x NVDLA v2.0 @ 614 MHz PVA v2 vision accelerator 100 TOPS AI performance (sparse) Video Encoder (H.265) 1x 4Kp60 | 3x 4Kp30 6x […]

January 10, 2023January 10, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

MistySoM Renesas RZ/G2L or RZ/V2L SoM and devkit goes for $112 and up

MistySOM devkit Renesas RZ/G2L or RZ/V2L

We’ve previously written about several system-on-modules and SBCs based on Renesas RZ/G2L or RZ/V2L Cortex-A55/M33 processors such as Geniatech “AHAURA” RS-G2L100 and “AKITIO” RS-V2L100 single board computers, Forlinx FET-G2LD-C system-on-module, and SolidRun RZ/G2LC SOM and devkit. But most of those are hard to buy, and you need to contact the company, discuss your project, etc… before purchase, except for the SolidRun Renesas RZ/G2LC Evaluation Kit going for $249. Another option is the MistyWest MistySOM module offered for $112 and up on GroupGets with either Renesas RZ/G2L or RZ/V2L processor, as well as an optional carrier board. MistySOM system-on-module MistySOM-G2L (aka MW-G2L) and MistySOM-V2L (aka MW-V2L) specifications: SoC – Renesas RZ/G2L or RZ/V2L with dual-core Cortex-A55 processor @ 1.2 GHz, Arm Cortex-M33 core @ up to 200 MHz, Arm Mali-G31 GPU, and DRP-AI vision accelerator (RZ/V2L only) System Memory – 2GB LPDDR4/DDR4 Storage – 32GB eMMC flash 2x 120-pin high-speed mezzanine […]

January 6, 2023 by Jean-Luc Aufranc (CNXSoft) - No Comments

Orbbec Femto Mega 3D depth and 4K RGB camera features NVIDIA Jetson Nano, Microsoft ToF technology

Orbbec Femto Mega is a programmable multi-mode 3D depth and RGB camera based on NVIDIA Jetson Nano system-on-module and based on Microsoft ToF technology found in Hololens and Azure Kinect DevKit. As an upgrade of the earlier Orbbec Femto, the camera is equipped with a 1MP depth camera with a 120 degrees field of view and a range of 0.25m to 5.5m as well as a 4K RGB camera, and enables real-time streaming of processed images over Ethernet or USB. Orbbec Femto Mega specifications: SoM – NVIDIA Jetson Nano system-on-module Cameras 1MP depth camera Precision: ≤17mm Accuracy: < 11 mm + 0.1% distance NFoV unbinned & binned：H 75°V 65° WFoV unbinned & binned：H 120°V 120° Resolutions & framerates NFoV unbinned: 640×576 @ 5/15/25/30fps NFoV binned: 320×288 @ 5/15/25/30fps WFoV unbinned: 1024×1024 @ 5/15fps NFoV binned: 512×512 @ 5/15/25/30fps 4K RGB camera FOV – H: 80°, V: 51°b D: 89°±2° Resolutions […]

January 5, 2023 by Jean-Luc Aufranc (CNXSoft) - 2 Comments

NXP i.MX 95 processor features Cortex-A55, Cortex-M33, and Cortex-M7 cores, eIQ Neutron NPU

NXP i.MX 95 is an upcoming Arm processor family for automotive, industrial, and IoT applications with up to six Cortex-A55 application cores, a Cortex-M33 safety core, a Cortex-M7 real-time core, and NXP eIQ Neutron Neural Network Accelerator (NPU). We’re just only starting to see NXP i.MX 93 modules from companies like iWave Systems and Forlinx, but NXP is already working on its second i.MX 9 processor family with the i.MX 95 application processor family equipped with a higher number of Cortex-A55 cores, an Arm Mali 3D GPU, NXP SafeAssure functional safety, 10GbE, support for TSN, and the company’s eIQ Neutron Neural Processing Unit (NPU) to enable machine learning applications. NXP i.MX 95 specifications: CPU Up to 6x Arm Cortex-A55 cores with 32KB I-cache, 32KB D-cache, 64KB L2 cache, 512KB L3 cache with ECC 1x Arm Corex-M7 real-time core with 32KB I-cache, 32KB D-cache, 512KB TCM with ECC 1x Arm Cortex-M33 […]

January 4, 2023 by Jean-Luc Aufranc (CNXSoft) - 7 Comments

Forlinx FET-MX9352-C – An NXP i.MX 9352 system-on-module for industrial AIoT applications

Forlinx FET-MX9352-C is a system-on-module based on NXP i.MX 9352 dual Cortex-A55 processor with Cortex-M33 real-time core and a 0.5 TOPS AI accelerator that can be used for industrial control, IoT gateways, medical equipment, and various applications requiring machine learning acceleration. The FET-MX9352-C follows last week’s announcement of the iWave Systems iW-RainboW-G50M OSM module and SBC with a choice of NXP i.MX 93 processors. The Forlinx module comes with two board-to-board connectors instead of solderable pads and can be found in the OK-MX9352-C single board computer with dual GbE, various display and camera interfaces, RS485 and CAN Bus, etc… FET-MX9352-C i.MX 9352 system-on-module Specifications: SoC – NXP i.MX 9352 with 2x Arm Cortex-A55 cores @ up to 1.7GHz (commercial) or 1.5 GHz (industrial), Cortex-M33 real-time core @ 250 MHz, 0.5 TOPS Arm Ethos U65 microNPU System Memory – 1GB/2GB LPDDR4 RAM Storage – 8GB eMMC flash 2x high-density 100-pin board-to-board […]

January 3, 2023 by Jean-Luc Aufranc (CNXSoft) - 1 Comment

MediaTek Genio 700 Cortex-A78/A55 IoT processor targets industrial and Smart Home applications

MediaTek Genio 700 is an octa-core Arm processor with two Cortex-A78 cores, six Cortex-A55 cores, a Mali-G57 GPU, and a 4 TOPS AI accelerator designed for consumer and industrial IoT applications. The new processor is a cost-down version of the Genio 1200 premium AIoT processor introduced last year with four Cortex-A78 and four Cortex-A55 cores. The Genio 700 offers many of the same features but with lower performance/capabilities, including a 3-core GPU and an AI accelerator limited to 4.0 TOPS, as well as support for dual displays up to 4K + Full HD (instead of 2x 4K), and 32MP single cameras (instead of 48 MP). MediaTek Genio 700 specifications: CPU – Octa-core processor with 2x Cortex-A78 cores @ up to 2.2 GHz, 6x Cortex-A55 cores @ up to 2.0 GHz GPU – Arm Mali-G57 MC3 GPU VPU Encoding up to 4Kp60 with H.265/HEVC Decoding up to 4Kp75, AV1, VP9, HEVC, […]

January 2, 2023January 7, 2023 by Jean-Luc Aufranc (CNXSoft) - 14 Comments

$150 Axelera M.2 AI accelerator module claims to deliver up to 214 TOPS

Axelera M.2 AI accelerator module is said to deliver up to 214 TOPS of AI inference and up to 3200 FPS with ResNet -50 in a compact M.2 2280 form factor. Few details are available at this time, but the module is based on the company’s Metis AIPU (AI Processing Unit) using in-memory computing based on arrays of SRAM memory devices used to “store a matrix and perform matrix-vector multiplications “in-place” without intermediate movement of data”. This technology is said to “radically” increase the number of operations per computer cycle with without suffering from issues such as noise or lower accuracy. The Metis AI platform delivers 50+ TOPS per core (RISC-V-controlled dataflow engine), offers FP32 equivalent accuracy, and has a 15 TOPS/W energy efficiency. The last point is impressive, but that means 214 TOPS won’t be reachable with the module shown above, since the M.2 form factor is designed to […]