In the context of its ecosystem, Xilinx will demonstrate solutions in the embedded, ML/AI, automotive and industrial sectors, using disruptive technologies designed for fast and accurate deployment by users. The following demonstrations are planned:
- 8-Channel Video Surveillance and Analytics – enables systems to simultaneously decode eight H.264/H.265 video channels from different cameras using flexible and responsive AI with a single Zynq UltraScale+ MPSoC.
- AWS IoT and Xilinx mobilize machine learning inference applications from cloud to edge with an oil and gas industry scenario. The demo shows how to enable an evolutionary application for durable systems by mobilizing a machine learning application from the cloud to the edge, resulting in lower latency and lower bandwidth requirements.
- SPYN demonstrates Python-based predictive maintenance – with a kit for the next generation of easy-to-use motor controls and analytics. Using the Zynq-7000 SoC, SPYN provides a C interface for creating accelerated motor control algorithms in an FPGA environment and the PYNQ framework for acquiring and processing sensor data using popular Python libraries.
- IDS NXT Vision app-based camera with Kortiq AIScale Machine Learning Accelerator – provides a range of objects that can be detected and detected in tens of milliseconds due to the low latency and low power consumption of Kortiq AIScale accelerators implemented in FPGA logic.
- Zynq UltraScale+ RFSoC Software Defined Radio – this product family provides a continuous RF signal chain for radio and cable access, test and measurement, early warning radar systems and other high-performance RF applications.
- Low Latency Multi-Stream Video Codec Compliant with H.264 and H.265 – This demo shows a low latency video codec in sub-frame latency mode supporting a multi-stream scenario that is ideal for live video streaming applications, including vision-based surgical equipment, surveillance cameras and ADAS.
- DesignStart FPGA – Cortex-M1 – enables hardware and software deployment on Xilinx devices with Arm Cortex-M1 core, as part of the Arm DesignStart FPGA program.
- DRIVE-XA Automated Driving Development Platform – a new development platform for automated driving systems and pathfinding. It offers users a modularly specifiable, heterogeneous processor environment with numerous options for sensor interfaces. The demo demonstrates the platform’s data aggregation, pre-processing and distribution (DAPD) functionality. It works with Xylon’s six-camera BSP framework, enabling users to efficiently start their own IP development.
- DeePhi LSTM speech recognition – an automatic speech recognition engine (ASR) based on the Xilinx FPGA developed for Deep Neural Networks (specifically LSTM)
- At the fair, Xilinx also introduces Alveo – designed to meet the ever-increasing demands of today’s data centers. The demonstrations also include a Video ABR transcoding application running on Alveo and an Xelera application showing acceleration with Alveo.
More information: www.xilinx.com or Embedded World Hall 3A, stand 235