NVIDIA cuVS is an open-source library designed to accelerate and optimize vector index builds and vector search for existing databases and vector search libraries. Built on top of the NVIDIA CUDA software stack, cuVS enables developers to enhance data mining and semantic search workloads, such as recommender systems and retrieval-augmented generation (RAG). It supports scalable data analysis, enhances semantic search efficiency, and helps developers accelerate existing systems or compose new ones from the ground up.

Features

• GPU-Accelerated Indexing Algorithms: Optimized GPU indexing enables high-quality index builds and low-latency search. cuVS delivers advanced algorithms for indexing vector embeddings, including exact, tree-based, and graph-based indexes.
• Real-Time Updates for Large Language Models (LLMs): Enables real-time updates to search indexes by dynamically integrating new embeddings and data without rebuilding the entire index. By integrating cuVS with LLMs, search results remain fresh and relevant.
• High-Efficiency Indexing: GPU indexing lowers cost compared to CPU-only workflows while maintaining quality at scale. Additionally, the ability to build large indexes out-of-core enables more flexible GPU selection and ultimately lower costs per gigabyte.
• Scalable Index Building: For real-time applications and large-scale deployments, cuVS enables both scale-up and scale-out for index creation and search at a fraction of the time it takes on a CPU without compromising quality.
• GPU-Accelerated Search Algorithms: Transforms vector search by integrating optimized CUDA-based algorithms for approximate nearest neighbors and clustering, ideal for large-scale, time-sensitive workloads.
• Low-Latency Performance: Provides ultra-fast response times for applications such as semantic search, where speed and accuracy are critical. Furthermore, support for binary, 8-, 16-, and 32-bit types means memory use is optimized for high-throughput applications.
• High-Throughput Processing: GPUs handle hundreds of thousands of queries per second, making cuVS perfect for demanding use cases like machine learning, data mining, and real-time analytics.

Benefits

• Enhanced Performance: Delivers significant performance improvements, making vector search and data clustering faster and more efficient.
• Increased Productivity: Reduces the time required to build and update indexes, allowing for higher throughput and faster turnaround times.
• Cost Efficiency: Reduces the number of systems required, leading to substantial cost savings in power and space.
• Scalability: Supports large-scale vector search and data clustering, making it suitable for complex and time-sensitive workloads.
• Versatility: Applicable to various data mining and semantic search applications, providing a comprehensive solution for developers.