Does the industrial AI edge gateway support collaborative inference between edge and cloud AI models?

The industrial AI edge gateway supports edge-cloud collaborative inference. Edge execution includes preprocessing and target detection, while cloud-based operation enables large-scale model fine-grained classification. Collaborative inference requires real-time network communication and thus high surge protection for network ports. It employs a common-mode choke CMZ2012A-900T and a TVS ESDLC3V3D3B, with two-stage protection added to the power input: a varistor 20D391K+TVS 5.0SMDJ33CA. Simultaneously, when downloading models from the cloud, an ESD0524P is applied to the storage interface. Measured collaborative inference latency is <200ms, with ESD ±8kV and no interruption.