Advantech Launches VEGA-340 Powerful Edge AI Acceleration Module for Versatile AI and Vision Analytics Applications
Taipei, Taiwan, March 2020 – Advantech (stock code: 2395), a leading global provider of IoT solutions, launches the VEGA-340 PCIe x4 Edge AI acceleration module for its VEGA-300 product line. With 8x Intel® Movidius™ Myriad™ X VPU on board, VEGA-340 provides substantial visual computing power to meet AI requirements for complex computing tasks such as in medical and AOI applications. Advantech provides an Edge AI Suite software toolkit which incorporates the Intel® Distribution of OpenVINO™ and provides a user-friendly graphical interface. Customers can use VEGA-340 and Edge AI Suite to conduct AI inference on edge devices using built-in pre-trained models or their own models while monitoring device status in real time.
Accelerating Edge AI Inference with High Power Efficiency
VEGA-340 offers a plug-and-play AI inference capability with a PCIe x4 form factor and 4x or 8x Intel® Movidius™ Myriad™ X VPUs on board. With the hardware accelerator components optimized for image processing, computer vision, and deep learning, VEGA-340 is ideal for AI-based vision applications such as machine vision, medical diagnostics, and video surveillance. Compared to the previous model VEGA-320 with 1x VPU, VEGA-340 is equipped with 8x VPUs that provide 8-fold performance. For object detection applications, running the MobileNet SSD model on an 8x Intel® Movidius™ Myriad™ X VPU is about six times more efficient than on an Intel® Core™ i3 processor. Customers can quickly integrate a VEGA-340 module with their existing equipment to conduct AI inference. They can enjoy the advantages of superior power efficiency and extra CPU/GPU resources, which enable additional application possibilities.
Edge AI Suite: User-Friendly Toolkit for AI Inference on Edge Devices
Advantech VEGA-340 comes with Edge AI Suite that integrates the Intel® OpenVINO™ toolkit and provides an easy to use graphical user interface (GUI) to enable AI inference on edge devices. Customers can run AI inference with the built-in pre-trained models or their own uniquely trained models. In addition to conducting AI inference on edge devices through the GUI, customers can monitor equipment status on the GUI dashboard. The latest version of Advantech’s Edge AI Suite integrates the Intel® OpenVINO™ Toolkit R3.1 and includes a new pre-trained model for accelerated YOLOv3 object detection that performs heterogeneous executions across the CPU/VPU to optimize workloads.
Advantech’s VEGA-340 Edge AI acceleration module is available now. For more information about these or any other Advantech products or services, contact your local sales representative or visit the Advantech website at https://www.advantech.com/products/edge-ai-acceleration-modules/sub_3d060f1e-e73e-460d-b38c-c69f76312c91