Please Use Streaming Workload to Benchmark Vector Databases

Why static workload is insufficient and what I learned by comparing HNSWLIB and DiskANN using streaming workload

Vector databases are built for high-dimensional vector retrieval. Today, many vectors are embeddings generated by deep neural nets like GPTs and CLIP to represent data points such as pieces of text, images, or audio tracks. Embeddings are used in many applications like search engines, recommendation systems, and chatbots. You can index embeddings in a vector database, which uses an Approximate …

audio benchmark clip dalle data databases data engineering embeddings generated gpts images neural nets retrieval streaming text vector vector database vector databases vectors vector search