Building AI Applications with GPU Vector Databases

In 2024 I was leading a workshop at the NVIDIA GTC (GPU Technology Conference) in San Jose, California. The workshop was held as a part of NVIDIA DLI (Deep Learning Institute) training sessions.

Abstract

The new wave of interest in large language models (LLMs) brought vector databases to the forefront of technologies for AI applications. We’ll clearly need accurate and performant query results from vector databases to meet an ever-growing demand for data and sophisticated applications, which is where GPUs offer their best. We’ll use a GPU-native vector database to introduce the concept of vector search, and work through use cases of vector search in practice. Specifically, we’ll use the vector database to build two AI applications that rely vector search: a document classification application and a LLM digital assistant chatbot.

Read more at NVIDIA website.