Tokyo: Kioxia Corporation has announced the integration of its KIOXIA AiSAQ, an Approximate Nearest Neighbor Search (ANNS) software technology, into the open-source vector database Milvus starting from version 2.6.4. This integration enables Milvus users to leverage KIOXIA AiSAQS's SSD-optimized vector search capabilities. It offers developers and enterprises a practical, cost-efficient approach to scaling AI applications without the typical challenge of increasing DRAM memory size associated with large-scale vector searches.

According to BERNAMA News Agency, the AI industry is transitioning from constructing massive foundation models to deploying scalable and cost-effective inference solutions that tackle real-world challenges. Central to this transition is Retrieval Augmented Generation (RAG), and KIOXIA AiSAQS technology has been developed to assist the community in utilizing SSD-based vector architectures. Its integration into the Milvus ecosystem facilitates easier adoption within the open-source community and aids developers in building faster and more efficient AI applications.