038-039_EETE-VF

EETE APRIL 2013

contact@absopulse.ch www.absopulse.ch High-Performance, high reliable Power Supplies 15 W to 12 kW+ adapted to your challenging projects l AC/DC Power Supplies l Rectifiers and Battery Chargers l DC/DC Converters l Sine-Wave Inverters l Phase and Frequency Converters l AC- Input, DC-output UPS systems l Open Frame l Wall mount l 19” Cassettes l DC-input dimmable back-light inverters for LCD displays l Complete Power Systems in 19" and 23" Racks ABSOPULSE Marketing Europe GmbH PoBox 1501 8620 Wetzikon CH Tel. +41 78 896 50 49 Fax +41 44 944 38 44 Write us: contact@absopulse.ch See us: www.absopulse.ch Fig. 2: Synapticon’s development environment for its Dynarc technology. “Our collaboration with Synapticon offers the service robotics industry a new way of solving problems,” said Ali Dixon, Director of Product Management and co-Founder, XMOS. “xCORE multicore microcontrollers offer deterministic performance via a fully-featured C programming environment that includes everything a software designer would expect from a best-inclass tool set. For users this brings reduced complexity and the flexibility of a programmable platform, while delivering hardware levels of responsiveness and latency.” Synapticon’s DYNARC Distributed Computing Technology is specifically designed for multicore devices such as xCORE. It enables software developers to easily configure and program individual distributed computing systems while using the computational resources of reconfigurable devices. It uses a high level model-based development environment that is combined with service-oriented architecture, code libraries and a compilation tool chain that makes it possible to easily run code and access peripheral resources anywhere in a network connecting 2 to 65k processor cores. Multiple sensor and actuator control signals have to be processed at the same time, occupying one or more conventional embedded computer only with signal processing and control loops, leaving no more resources for interpretation and cognition tasks. Often there are long distances between the peripheral devices and the core computing system, resulting in latency issues for high-speed control applications, even given current real-time communication technologies. Developers thus need to build control networks, where a major part of the data is processed by algorithms right at the peripheral devices. Classic smart sensor or actuator concepts are not a viable solution here, as they usually cannot execute individual code and provide no options for sensor signal fusion or parallel use of sensor signals, which is crucial for most Cyber-Physical Systems. The DYNARC hardware architecture and middleware has been developed to provide distributed real-time control and data acquisition applications. Computing nodes based on DYNARC Stream technology provide both resources for multiple tasks as well as specific computing resources for each individual algorithm. Using an FPGA and a multicore FPU as coprocessor, DYNARC Stream technology provides developers with a fully adaptive high-end computing machine that perfectly fits the demand of the application they develop, while consuming a fraction of energy compared with high-performance CPUs. This creates an autonomous system that is able to adapt its resources to the situation at hand, similar to the way humans concentrate on a certain task. www.electronics-eetimes.com Electronic Engineering Times Europe April 2013 37


EETE APRIL 2013
To see the actual publication please follow the link above