Zilliz's paper selected for the top database conference VLDB'22

Recently, the latest database research results of the Zilliz technical team have been recognized internationally. The cloud native database system Milvus 2.0 paper "Manu: A Cloud Native Vector Database Management System" designed for vector data management was selected into the top database conference VLDB'22 .

As a start-up company, Zilliz has always been a leader in research and innovation in the database field. Previously, Milvus' paper "Milvus: A Purpose-Built Vector Data Management System" was included in the top database meeting SIGMOD'21. In the BigANN vector retrieval challenge of NeurIPS'21, the world's top academic conference in the field of artificial intelligence, Zilliz won the championship with its self-developed disk storage optimization algorithm, further expanding the boundaries of vector search algorithms.

VLDB (International Conference on Very Large Databases) is an annual international conference widely attended by database researchers, manufacturers, application developers, and users. It, together with SIGMOD and ICDE, are recognized as the three top international academic conferences in the field of data management and databases. The included papers represent the highest level in the field of database technology, and are also an important indicator of future technological development.

Manu: A Cloud Native Vector Database Management System

"Manu: A Cloud Native Vector Database Management System" introduces in detail the actual needs of users for vector data management and the architecture design of a new generation of distributed cloud-native vector databases . Milvus 2.0 creatively uses logs as the system backbone to connect various components in the system, which provides greater convenience for the independent elasticity of each component, functional evolution, and isolation of resources and faults, and achieves dynamically adjustable consistency based on logs and time semantics A flexible solution that gives users more freedom to choose between consistency, cost, and performance.

The VLDB judges spoke highly of the cutting-edge and innovative nature of Milvus 2.0: "The distributed cloud-native vector database is a new bright spot in the database field. Milvus 2.0 related papers expounded the design concept and practical application of the cloud-native vector data management system. Case study, and discussed the future development direction of vector database, which has great inspiration and reference significance for the development and evolution of database in the industry."

Star Lord, founder and CEO of Zilliz, said: "Zilliz has attached great importance to the investment in cutting-edge technology research from the first day of its establishment, and has established a special research department. We will apply the latest research results to the development and development of Milvus products. In the process of iteration, we will continue to create value for users. The selection of VLDB'22 represents the industry's re-recognition of Zilliz's technological leadership. In the future, we will continue to seek technological innovation breakthroughs and work with research institutions, community users and ecological partners to promote the latest The practice of vector database research results in AI scenarios.”

Reference link

Manu: A Cloud Native Vector Database Management System, VLDB 2022

Milvus: A Purpose-Built Vector Data Management System, SIGMOD 2021

Billion-Scale Approximate Nearest Neighbor Search Challenge: NeurIPS'21 competition track

About Milvus

Milvus is currently the most active and popular open source vector database project in the world, with more than 12k GitHub Stars. After continuous polishing by the technical community and industry users, Milvus has gradually formed a set of best practices for large-scale distributed vector databases. At present, Milvus has been trusted by more than 1,000 corporate users around the world, and is widely used in recommendation systems, video analysis, image retrieval, text retrieval, cross-modal retrieval, targeted advertising, personalized search, intelligent customer service, fraud detection, network security and new drug discovery.

GitHub: https://github.com/milvus-io/milvus

Docs: https://milvus.io/docs

Official website: https://milvus.io/

---

Zilliz is a pioneer and global leader in the field of vector database systems, developing vector database systems for AI production systems. With the mission of exploring the value of unstructured data, Zilliz is committed to creating a new generation of database technology for AI applications, helping enterprises to easily develop AI applications. Zilliz's products can significantly reduce the cost of managing AI data infrastructure, helping AI technology empower more businesses, organizations and individuals.

For event information, technology sharing and recruitment express, please follow: https://zilliz.gitee.io/welcome/