DFI is devoted to developing AI edge computing products. Embedded solutions effectively reduce basic equipment and utilize more existing resources; furthermore, they also provide a highly flexible and scalable platform to meet the needs of various workloads, facilitating the integration of different devices and simplifying the architecture. DFI has taken the lead in jointly integrating ultra-compact embedded products, and Intel display chip processors to promote iGPU (Integrated Graphics) SR-IOV virtualization technology and commercialize modules.
With the current Software-Defined AIoT trend, hardware-supported virtualization technology is essential in the application environment. DFI products ranging from the industrial grade motherboard, ADS310, equipped with 12th generation Intel CPU, to the EC70A-TGU system and tablets, leverage virtualization technology to integrate and set up multiple virtual platform spaces. In addition to optimizing operations, it saves the cost of connecting data to the cloud and improves yield and efficiency.
DFI President Alexander Su said, “Benefiting from Intel’s technology and the support of SR-IOV architecture, the hardware can execute different operating systems through a single CPU and integrate large amounts of shared data. Due to substantial improvement in drawing performance within a virtual environment, features such as color recognition, measurements, and appearance flaw detection have increased in speed.”
Intel believes that cost considerations are a vital concern in industrial automation. DFI’s hardware can support SR-IOV virtualization technology. Through workload consolidation, it can improve cost efficiency, resolve the long term problem of graphics computing performance, and fulfill the needs of large-scale deployments in application fields.
DFI’s ECI and EIAMR products have been tested by Intel’s ESDQ, and DFI and Intel have released a technical white paper. Through working closely with Intel, DFI continues to assist and optimize the development of various services in IoT applications, accelerating the realization of Software-Defined IoT.
Smart poles and smart retail are also gradually moving toward the trend of workload integration. Smart Poles can identify pedestrian behavior with roadside and vehicle equipment and synchronize the data to the smart poles. Many devices can also execute high-performance AI edge computing and speed up people flow detection, people counting, monitoring, etc. Regarding smart retail, identification systems can be implemented to determine consumer age and gender to understand customers.