IoT Archives - Page 344 of 453 - SmartCitiesElectronics.com

Arm introduces CPU, GPU, NPU and customisation programme

Wed, May 27, 2020

Multiple announcements by Arm form its next-generation mobile offering, with a new Cortex-A78 CPU, a new GPU on the Valhall architecture, an enhanced neural processing unit and a customisation programme.

For PC-level productivity in smartphones, the Cortex-A78 CPU is described by Arm as it is efficient Cortex-A CPU ever designed for mobile. It represents a 20 per cent increase in performance compared with Cortex-A77-based devices, in anticipation of 5G services. It also has a more efficient management of compute workloads and greater on-device machine learning (ML) performance, added Arm. The Cortex-A78’s performance-per-Watt makes it suitable for computing on foldable devices with multiple and larger screens.

Following last year’s introduction of the Mali-G77 GPU on the Valhall architecture, Arm has also increase graphics performance, with the Mali-G78, which delivers a 25 per cent increase compared to the Mali-G77. Asynchronous top level, tiler enhancements and improved fragment dependency tracking have resulted in support for up to 24 cores. Additionally, power- and energy-efficiency contributes to extended mobile device battery life, for users to enjoy mobile entertainment experiences for even longer while on the go. For developers, this means content can be easily optimised to run seamlessly on Arm Mali GPUs, confirmed the company. Enhanced tools include the Performance Advisor, which allows quick detection of bottlenecks and real-time reporting to enable continuous integration and faster workflow.

The premium Mali-G78 is accompanied by the Arm Mali-G68, which supports up to six cores.

To address expanding machine learning (ML) use cases from augmented reality (AR) -based smartphone applications to smart home hubs, the Ethos-N78 neural processing unit (NPU) has been introduced. It offers 25 per cent more performance efficiency than the Ethos-N77, and delivers greater on-device ML capabilities, said Arm. The Ethos-N78 is claimed to offer unprecedented levels of configurability with available configurations starting at one up to 10 Terra operations per second (TOPS).

Arm also introduced the Cortex-X Custom program for customisation and differentiation beyond the traditional roadmap of Arm Cortex products. The programme’s first CPU is the Cortex-X1 – the most powerful Cortex CPU to date. It has a 30 per cent peak performance increase over Cortex-A77.

http://www.arm.com

> Read More

ToF sensor calculate range of multiple objects

Wed, May 27, 2020

STMicroelectronics has introduced a time of flight (ToF) sensor that enables multi-object ranging. The VL53L3CX can be used for industrial and personal electronics applications. It is the latest addition to the FlightSense sensors range and has patented histogram algorithms that allow measuring distances to multiple objects. It is also claimed to increase accuracy.

The VL53L3CX measures object ranges from 25mm to three metres and is unaffected by the target colour or reflection levels, unlike conventional infra red sensors, pointed out STMicroelectronics. As a result, designers can introduce product features. For example, occupancy detectors can be enabled to provide error-free sensing by ignoring unwanted background or foreground objects. They can also report the exact distances to multiple targets within the sensor’s field of view.

The patented histogram algorithms increase cover-glass crosstalk immunity and allow real time smudge compensation preventing external contamination from adversely affecting the ranging accuracy of equipment that may be used in a dusty industrial environment. Ranging under ambient lighting is also improved, noted STMicro.

Superior linearity is claimed to increase short-distance measurement accuracy enhancing wall tracking, faster cliff detection and obstacle avoidance in equipment such as service robots.

In common with the existing FlightSense sensors, the VL53L3CX has a compact package design for integration into customer devices, and low power consumption to extend battery runtime.

The VL53L3CX can be used in a variety of applications, including occupancy detection in building automation and lighting controllers, smarter proximity sensing in IoT endpoints, auto-wakeup in portable devices and robust user detection in automatic sanitary equipment. The accuracy and response times also make them suitable for devices requiring precise movement control, such as robotics and indoor drones.

The VL53L3CX is available now.

http://www.st.com

> Read More

Aaeon Ups its game for intelligence surveillance

Wed, May 27, 2020

To increase the capability of surveillance systems with the use of intelligent edge computing, Aaeon’s UP Xtreme Smart Surveillance kit provides the technology needed to power smart surveillance systems, such as surveillance cameras that analyse traffic using heat maps or automated control.

Smart surveillance adds intelligent processing of data that is already being gathered to analyse traffic with heat maps, monitor areas with virtual fences, or react to data in other ways to allow for automated control or alerting staff on premises.

Aaeon’s partners include Intel, Milestone and Saimos. The kit works as an AI-enabled network video recorder (NVR), for example to monitor video streams supporting up to 32 cameras simultaneously, using the Intel Core i7 model with Milestone XProtect Express+, explained Aaeon.

UP Xtreme Smart Surveillance integrates video management software (VMS) from Milestone with video analytic software from Saimos, then adding a deep learning AI edge inference through the Intel Distribution of OpenVINO toolkit. The combined package saves deployment time and cost for end users, as well as overcoming challenges such as storage issues, said Aaeon. The AI inference is also accelerated with two Intel Movidius Myriad X VPUs, helping provide even more power alongside the 8th Generation Intel Core processor supplied with the UP Xtreme system.

“UP Xtreme Smart Surveillance offers video analytics combined with a modern video management system, integrating everything into a one-box solutions which can be deployed with both new or existing infrastructure,” said Jürgen Konetschnig, CTO of Saimos. “The hardware and software are integrated to make deployment of smart surveillance applications easier and simpler for users in any field,” he added.

The Boxer-8130AI can be integrated into any environment, said Aaeon. It has six MIPI CSI-2 interfaces, to support up to six MIPI cameras. For easy maintenance it has an SD Card slot, USB OTG, remote on/off and almost all of its I/O ports together on one side. The Boxer-8130AI also features two antenna ports, for mobile or AI of thing (AIoT) gateway applications.

Once installed, UP Xtreme Smart Surveillance provides intuitive analytic parameter settings to provide a comprehensive overview of the surveillance system, said the company. UP Xtreme Smart Surveillance supports people or object counting, heat mapping, virtual fence and perimeter protection, dynamic blurring and object detection functions. It uses data analytics and edge computing to provide real time analysis, and can provide proactive measures through alerting relevant staff as soon as an incident is detected.

UP Xtreme Smart Surveillance is scalable in terms of both the number of cameras to AI acceleration.

http://www.aaeon.com.

> Read More

AI stack doubles performance for edge AI

Fri, May 22, 2020

For on-device AI processing at the edge, Lattice Semiconductor has introduced the sensAI 3.0, available on low power, 28nm FD-SOI-based Lattice CrossLink-NX FPGAs.

This latest version of the company’s stack features customised convolutional neural network (CNN) IP. This flexible accelerator IP simplifies the implementation of common CNN networks, explained Lattice, and is optimised to leverage the parallel processing capabilities of FPGAs. Adding support for CrossLink-NX FPGAs brings new levels of power and performance to smart vision applications in the surveillance/security, robotics, automotive and computing markets, said the company.

To address data security, latency and privacy issues, developers want to move the AI processing that powers their smart vision and other AI applications from the cloud to the edge. Most edge devices are battery-powered or sensitive to power consumption, so developers need hardware and software that deliver the processing capabilities needed for AI applications, while keeping power consumption as low as possible.

For applications like smart vision that require higher edge AI performance, CrossLink-NX FPGAs running sensAI software deliver twice the performance at half the power when compared to prior releases of the solutions stack, confirmed Lattice.

Updates to the NN compiler software tool let developers easily compile a trained NN model and download it to a CrossLink-NX FPGA.

A new feature for this version is a VGG-based object counting demo operating on a CrossLink-NX FPGA. It delivers 10 frames per second while consuming only 200mW. Object counting is used in smart vision applications that are used in the surveillance, security, industrial, automotive and robotics markets.

When running on a CrossLink-NX FPGA, the sensAI solutions stack offers up to 2.5Mbit of distributed memory and block RAM with additional DSP resources for efficient on-chip implementation of AI workloads, this reduces the need for cloud-based analytics.

Being manufactured in a 28nm FD-SOI process means that the CrossLink-NX FPGAs deliver a 75 per cent reduction in power, in comparison to similar, competing FPGAs, claims Lattice.

A target application for sensAI is smart vision, and CrossLink-NX devices are currently the only low-power FPGAs to deliver MIPI I/O speeds of up to 2.5Gbits per second to support components such as image sensors. This makes CrossLink-NX FPGAs a suitable hardware platform for sensAI applications requiring MIPI support. CrossLink-NX FPGA’s I/Os offer instant-on performance and are able to configure themselves in less than three milliseconds, with full-device configuration in as little as eight milliseconds.

There is also increased neural network architecture support in the form of support for the MobileNet v2, SSD, and ResNet models on the Lattice ECP5 family of general-purpose FPGAs.

Lattice Semiconductor specialises in low power programmable devices, working with customers across the network, from the edge to the cloud, in the communications, computing, industrial, automotive and consumer markets.

http://www.latticesemi.com

> Read More

Arm introduces CPU, GPU, NPU and customisation programme

ToF sensor calculate range of multiple objects

Aaeon Ups its game for intelligence surveillance

AI stack doubles performance for edge AI

About Smart Cities