Company
Date Published
Jan. 9, 2025
Author
Eric Hanson, Kristin Tufte
Word count
1223
Language
English
Hacker News points
None

Summary

The world has changed with the advent of ChatGPT, sparking a revolution in how we interact with AI. Large Language Models (LLMs) have ushered in a new era of applications like semantic search and Retrieval-Augmented Generation (RAG), which rely on vector search — a critical enabler for modern applications striving to deliver smarter, faster and more intuitive user experiences. Specialty databases like Pinecone and Zilliz have demonstrated the value of purpose-built vector databases in accelerating AI-driven workloads, while virtually all major SQL and NoSQL databases have responded by adding indexed vector search capabilities. However, achieving competitive queries per second (QPS) per dollar at a fixed level of recall is crucial for many applications, and SingleStore has validated that it delivers competitive QPS/$ for vector workloads while offering robust analytics and transactional capabilities. SingleStore's performance tests show cost-competitive performance with specialty vector databases like Pinecone and Zilliz, combining competitive vector search performance with fast SQL analytics, joins, and aggregations across petabytes of structured and semi-structured data to power intelligent applications.