NEAREST BY Join: Scaling Vector Search in Databricks Runtime
- Databricks: 95 events in the last 90 days
- Previous: earlier the same day · Unlocking Data Portability: Preventing Catalog Lock-in with REGISTER and UNREGISTER APIs
What happened
How Databricks built batch vector search as a native SQL join: NEAREST BY syntax, Photon GEMM kernels, and an IVF Vector Index built on Liquid Clustered Delta tables.
Summary assembled by rule from the sources below
Why it's spreading
Sources
Official
- NEAREST BY Join: Scaling Vector Search in Databricks RuntimeDatabricks Blog