Session

DescriptionThis session features seven papers focused on accelerating retrieval-augmented generation (RAG), approximate nearest neighbor search (ANNS), and graph-based vector queries through in-memory and near-memory designs. Emphasizing hardware-software co-optimizations, these contributions reduce latency in data-intensive search operations, such as vector pruning, graph traversal, and embedding handling, to support scalable AI applications like point cloud processing and billion-scale datasets.