Predicted growth for memory and visual AI at the edge

According to a DIGITIMES report, Novatek estimates that the demand for memory and visual artificial intelligence solutions at the edge will be one of the main drivers of industry growth by 2026. This forecast underscores the increasing importance of distributed processing and the integration of AI directly into edge devices.

The increased demand for memory is likely linked to the need to support increasingly complex AI models and large datasets, necessary for training and inference at the edge. Visual AI, in particular, requires high computing and memory capabilities for real-time image and video processing.

This trend could prompt companies to evaluate on-premise or hybrid solutions to manage AI workloads, balancing the benefits of low latency and data sovereignty with the scalability and flexibility of the cloud. For those evaluating on-premise deployments, there are trade-offs that AI-RADAR analyzes in detail on /llm-onpremise.