turbopuffer
@turbopuffer
{vector, full-text} search engine built on object storage. fast, cheap, trillion scale. powers Anthropic, Harvey, Notion, Cognition, and more
Joined October 2023
6 Following 17.8K Followers
rescoring fewer full-precision vectors cut disk read I/O by up to 80% on one prod cluster
lower disk I/O produces higher query throughput, especially on smaller VMs where full-precision vectors don't fit in memory
it also produces beautiful charts
Show more
binary quantization lets tpuf store & score more vectors in RAM. to keep recall high, we rescore full-precision candidates from SSD
we cut vector search latency up to 4.3x on low-mem VMs by dynamically pruning rescore candidates
less rescoring→ fewer SSD reads→ faster search
Show more
