Perform High-Efficiency Search, Improve Data Freshness, and Increase Recall With GPU-Accelerated Vector Search and RAG Workflows

, Principal Engineer, ML, Data Mining, and Vector Search, NVIDIA
, CEO, Zilliz
Vector databases are a foundational component of many AI applications. They assist in model training, facial recognition, recommendation systems, semantic understanding in large language models, and more. As the number of AI applications grows, vector databases are presented with more complicated use cases. These use cases put pressure on vector databases to provide low latency, high throughput, and high levels of recall while scaling up the size and quantity of embeddings. Milvus, one of the world's leading vector databases, uses GPUs to meet the growing demands of AI applications. Learn how the integration of the RAPIDS RAFT library significantly enhances Milvus's performance for building indexes and performing search queries. We'll discuss the architecture of the core integration and provide benchmarks in real-world applications.
活动: GTC 24
日期: March 2024
行业: 所有行业
话题: Data Analytics
NVIDIA 技术: Grace CPU,Hopper
级别: 中级技术
语言: 英语
所在地: