HAVEN applies HBF to billion-scale vector search

HAVEN proposes using HBF as a high-capacity tier for approximate nearest-neighbor search, moving beyond the dominant LLM-weight use case.

Why it matters

It broadens the application space and exposes workload-specific questions about access granularity, locality, and update behavior.

Exact source title

HAVEN: High-bandwidth flash memory Accelerated VEctor iNdex