VikingDB - High Performance Cloud Native Vector Database by Volcano Engine

What is VikingDB

VikingDB is a high-performance cloud-native vector database from Volcano Engine, designed for processing massive high-dimensional vector data.VikingDB has multiple data writing methods, including real-time synchronous, asynchronous writing, etc., to meet the needs of different scenarios.VikingDB is based on self-developed high-efficiency indexing algorithms, such as HNSW and IVF, to achieve tens of billion vectors and millisecond retrieval, and is also compatible with dense and sparse vector retrieval. VikingDB provides SaaS console, APIs and SDKs in multiple languages, and supports automatic elastic expansion, which can effectively reduce storage costs.VikingDB is widely used in the fields of multimodal search, intelligent recommendation, RAG scenarios, and memory construction, which can help enterprises realize efficient data management and intelligent application development.

Key Features of VikingDB

• Flexible data writing: Provide multiple write methods, such as real-time synchronous, asynchronous write, single data write and batch write, to meet the needs of different business scenarios.
• Efficient indexing and real-time updatesThe indexing algorithm is based on self-developed HNSW, IVF and other advanced indexing algorithms, combined with the streaming update architecture, to ensure that the data can still be updated quickly under high load, and to guarantee the real-time retrieval.
• Powerful search capabilities: It supports vector, scalar, mixed and multimodal data retrieval, and can achieve millisecond retrieval in tens of billions of vector data to meet complex query requirements.
• Elastic and scalable cloud servicesIt provides SaaS console, API interfaces and SDKs in multiple languages, supports automatic elastic expansion, and facilitates users to quickly build and manage data retrieval processes.
• High performance and cost optimization: Through deeply optimized indexing algorithms and quantization techniques, it achieves extremely high retrieval efficiency while reducing storage costs to ensure cost-effectiveness in large-scale data scenarios.
• Knowledge Base and Memory Functions: It supports complex semantic retrieval and long-term memory storage of large models, and is suitable for personalized interaction scenarios, such as intelligent assistants, education and teaching.

VikingDB's official website address

• Official website address:: https://www.volcengine.com/product/VikingDB

How to use VikingDB

• Register & Login: Visit the Volcano Engine website, complete account registration and login to access the VikingDB console.
• Creating Instances: Create a VikingDB instance in the console, configuring parameters such as name, storage capacity, and performance specifications as required.
• Data preparation and vectorization: Organize the data to be processed and use Embedding models (e.g. Doubao or other open source models) into vector form.
• Access to SDK: Install and initialize the SDK provided by VikingDB (supports Python, Java, Go, etc.) and connect to the created database instance.
• write data: Write vector data to VikingDB based on the SDK, choosing from real-time synchronous, asynchronous, and other write methods.
• retrieve: Use the SDK to perform vector, scalar, or hybrid searches to get the most similar results.
• Monitoring and Optimization: Monitor instance performance metrics on the console and adjust configurations to optimize performance and cost based on actual usage.

Core Benefits of VikingDB

• High Performance SearchBased on self-developed efficient indexing algorithms (e.g. HNSW, IVF, etc.), it realizes millisecond retrieval in tens of billions of vectors, and the retrieval latency is as low as less than 10ms, which significantly improves the query efficiency.
• Diversity data supportIt supports the retrieval of dense vectors and sparse vectors, and is compatible with the retrieval of vector, scalar, mixed, and multimodal data, and is applicable to a wide range of complex data types and scenarios.
• Flexible data writingIt provides real-time synchronization, asynchronous writing, single data writing and batch writing to meet the data writing requirements in different business scenarios and ensure the flexibility and efficiency of data processing.
• Elasticity and scalability: As a cloud-native database, it provides SaaS console, APIs and SDKs in multiple languages, supports automatic elastic expansion, and dynamically adjusts resources according to data volume and query load to ensure system stability and efficiency.
• Low-cost storage: Use deeply optimized indexing algorithms and quantization techniques to achieve high performance while reducing storage costs and improving price/performance.
• Knowledge Base and Memory Functions: Provide complex semantic retrieval and long-term memory storage of large models, applicable to personalized interaction scenarios, such as intelligent assistants, education and teaching, role-playing, etc., and support efficient retrieval and memory management of large models.

Who VikingDB is for

• Artificial Intelligence and Machine Learning Engineer: For engineers who need to process and retrieve large-scale vector data, efficient tools are provided to support model training, feature retrieval, and multimodal data processing.
• data scientist: In the process of data analysis and mining, diverse retrieval capabilities and flexible data writing can help data scientists quickly validate models and process complex data.
• Corporate Technical TeamVikingDB's high performance and elastic scalability supports business growth and technology needs for organizations that need to build intelligent recommendation systems, multimodal search platforms, or knowledge bases.
• system architect: For architects responsible for the design of system architectures, provide high-performance, scalable and easy-to-integrate solutions that fit seamlessly into existing technology stacks.
• developersVikingDB provides a multilingual SDK and detailed development documentation to facilitate developers to quickly access and use VikingDB for efficient data management and application development.