2023, Taipei, China – Advantech, a global provider of intelligent Internet of Things systems and embedded platforms, announced an industrial-grade GPU graphics solution developed in partnership with Intel using Intel Arc graphics cards. The collaboration aims to meet the market’s growing demand for GPU and visual AI performance, while developing a more open AI ecosystem that offers a wide range of products. Including the Advantech VEGA-P110, a PCIe GPU card; And Advantech VEGA-X110, an MXM Type A 3.1 GPU card that uses Intel ARC A730M Gpus in both specifications; The product line offers excellent performance and price while enhancing image processing and edge AI acceleration in medical imaging, gaming, and factory automation applications.

901E-2500-RBC

Fast integration of multiple applications

Advantech’s VEGA-P110 and VEGA-X110 solutions are developed using Intel Arc A370M Gpus. The GPU has eight Xe cores and 128 Intel Xe matrix extension engines, and supports H.264/H.265/AV1/VP9 codec technology. In particular, the VEGA-X110 MXM GPU card offers a wide voltage input of 9-20V for easy integration. The VEGA-P110 GPU provides Gen.4 PCIe x16 ports. Combined with excellent thermal design, both interfaces support integration with embedded computers, medical systems, and gaming platforms. And provides support for a variety of applications.

Ideal video codec, AI and graphics performance

The combination of GPU-Iris Xe and VEGA GPU cards integrated on the system improves video coding performance by 60% and improves AI-enhanced video upgrades by 2.91 times with XMX Hyper Compute acceleration enabled. In addition, both the VEGA-X110 and VEGA-P110 leverage Intel Deep Link to accelerate AI and graphics computing performance while optimizing CPU and GPU workloads. These solutions also leverage Intel OpenVINO, an open source toolkit containing more than 100 pre-trained models and popular public frameworks. Advantech VEGA GPU cards provide a more open and accessible AI development environment for developers seeking to optimize and deploy AI inference solutions.