New Upgrades to Zilliz Cloud, a look into Cardinal Vector Search Engine, Meetup recap, Intel and Milvus collaboration, and Valentine’s Day poems
In this issue:?
?? New Upgrades to Zilliz Cloud are here
In the past few weeks, Zilliz has introduced several new features to enable AI innovations with our vector database and data services.?
If you’re tired of inefficiently updating data in a database through a two-step process, try Zilliz’s Upsert function.
If you’re tired of recommendations that are too similar or disparate, try Zilliz’s Range Search feature.
If you use Google Cloud's resources and infrastructure, try accessing Zilliz Cloud through the GCP Marketplace.
If you have existing data on Spark, try migrating it to Zilliz Cloud using the Databricks Connector.
If you collaborate with multiple business units and manage a vector database for various AI applications, try granular Role-based access control.
If any of these sound like YOU ??, start using the features by signing up here.?
?? Try the new Cardinal Search Engine for Vector Search
Did you hear? There’s a new vector search engine and it has a remarkable 10x performance boost! Introducing Cardinal: the most performant engine for vector searches. Cardinal implements various algorithms-related, engineering, and low-level optimizations by introducing the AUTOINDEX mechanism. This automatically selects the best search strategy and index for the dataset, which eliminates the need for manual tuning.?
Learn more about Cardinal’s capabilities and compare it to other vector search engines in this blog.?
?? Community Spotlight: February Unstructured Data Meetups
?? How Milvus and Intel AVX-512 Increase Vector Search Performance?
Milvus, powered by Intel, excels in heterogeneous computing for Vector Search - a game-changer in AI applications like Chatbots and fraud detection, handling billion+ vectors efficiently. @ Intel Business and Zilliz ensure seamless hardware acceleration, empowering developers for high-performance AI solutions. The Milvus core search engine (Knowhere) uses AVX-512 for the most performance-critical code.
Performant indices like HNSW and IVFFlat saw a performance increase of about 10% utilizing AVX512, while IVFSQ saw 2x performance since memory bandwidth and compute is often the limiting factor.
Read more and see all the results in this blog by Zilliz and Intel!
领英推荐
?? Using GenAI to Write a Valentine’s Day Poem?
February is the month of love, so we asked you to write your best Valentine’s Day poems using ChatGPT. ??
Below is the winning poem! Feel free to generate and comment another GenAI-generated poem in this post — February is not over yet!
Upcoming Events?
Feb 22: Optimizing Your Milvus Instance (virtual)?
Xiaofan(James) Luan will share valuable insights and expert tips to unlock the full potential of Milvus. Discover advanced techniques, best practices, and performance-enhancing measures that will take your experience to the next level. This webinar is for ALL levels of experience, so don’t miss your chance to learn something new!?
Register here!
Feb 29: Advanced Retrieval Augmented Generation Apps with LlamaIndex (virtual)?
Join Laurie Voss from LlamaIndex for the basics of how RAG works under the hood AND advanced techniques for when you want to take your app further.?
Register here!
March 8: WiDS Stanford watch party (in-person)?
What better way to celebrate National Women’s Day than with a group of local women watching Stanford’s Women in Data Science Conference? Our DevRel, Christy Bergman is hosting a watch party in SF for the conference happening on March 8! Join her and other women in tech to listen to outstanding women discuss their work in Data Science, AI, and related fields. ??
Register here: https://lu.ma/9c5pat0i?
March 14: Choosing the Right Vector Database: A Practical Guide (virtual)
In this webinar, Li Liu will explain how to select the perfect vector database for your specific use case. He will break down the key factors to consider, demystify complex terminology, and provide actionable insights to help you make an informed decision.
Register here!
--
See you next time! ??