The BM25 guy on vector retrieval: scaling HNSW to billions of vectors in production
jobergum · x · 2026-11-20
jobergum of Vespa, known as "the BM25 guy", says embedding-based retrieval is here to stay: design partners run billions of vectors in production with ever-increasing QPS from agents — a workload where HNSW really shines.
The linked post by CTO Henning Baldersheim explains how they scale HNSW, and importantly, where HNSW is not a good fit.
Related event: Scaling HNSW to Billions of Vectors(2 posts)→
More from Infra
- AMD's Lisa Su: data center to hit ~80% of revenue by 2027, sees $220B server CPU market — AccBalanced · 2026-11-20
- llms.txt "never really existed": Mintlify CEO on agents now driving 65% of doc traffic — zeeg · 2026-11-20
- Ben Bajarin: $1.8T cloud backlog says the AI buildout is still early despite investor doubt — BenBajarin · 2026-11-20
- Quantized code LLMs measured by energy per solved task: 7B-q2_K costs 2x — Aditi Patodiya · 2026-11-20
- Kale TTS runs 65-180x realtime on laptop CPU, deploys to Raspberry Pi and phones — NERDDISCO · 2026-11-20
- Geoffrey Huntley's six takes: compute crunch lasts 6-10 years, open weights trail by 3 months — sebpaquet · 2026-11-20