|Zilliz Cloud Adds Range Search
|Written by Kay Ewbank
|Tuesday, 06 February 2024
The Zilliz Cloud database developers have added new features to Zilliz Cloud, including range search, multi-tenancy & RBAC, and up to ten times improved search and indexing performance.
Zilliz Cloud is built on the open-source Milvus vector database. Zilliz is a vector database platform that can be used to store, index and search for data within vectors. Zilliz was developed by part of the original team who developed the Milvus vector database.
Milvus was designed to store, index, and manage massive embedding vectors generated by deep neural networks and other machine learning (ML) models. Unlike existing relational databases which mainly deal with structured data following a pre-defined pattern, Milvus is designed from the bottom-up to handle embedding vectors converted from unstructured data.
Unstructured data is converted into vectors using embedding techniques, and Milvus (and Zilliz) are able to analyze the correlation between two vectors by calculating their similarity distance. If the two embedding vectors are very similar, it means that the original data sources are similar as well.
The improved performance of the new version of Zilliz Cloud compared to Milvus comes from its use of Hierarchical Navigable Small World (HNSW) graph indexes. Most vector databases now include support for HNSW, so the addition of this feature was seen as key for Zilliz.
The version of HNSW that has been included in Zilliz is Approximate Nearest Neighbor Search (ANNS). This is used in Zilliz Cardinal, a multi-threaded, high-efficiency vector search engine that works with various data formats, including FP32, FP16, and BF16.
Zilliz Cloud is also now available on Google Cloud Marketplace, alongside the existing support for AWS and Azure. The Zilliz team says the support for Google Cloud eliminates the need for upfront budget planning and commitments by letting users scale their clusters and pay as they need.
Other improvements to the new version include improved query precision and the addition of a Cosine Similarity metric, Range Search, and Upsert.
Zilliz Cloud comes with a unified Milvus client, and offers one-click data migration between Milvus and Zilliz Cloud. The new release also offers granular Role-Based Access Control (RBAC) and comprehensive API keys for improved security.
Zilliz Cloud is available now.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Tuesday, 06 February 2024 )