Embeddings & Vector SearchJeff Johnson, Matthijs Douze, Hervé Jégou (FAIR)FAIR

FAIR presents FAISS

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

Presented FAISS — GPU k-selection reaches up to 55% of theoretical peak. Handled 1 billion vectors on 4 Titan X GPUs in under 12 hours

Why it matters

The library that brought large-scale vector search into the realm of practical use

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/1702.08734

Added in Week 35, 2026 · See that week

On the map — ⑰ Embeddings line

  1. Previous stopword2vec
  2. This record · stopFAISS
  3. Next stopSentence-BERT
See it on the map

Nearby on the same axis

Malkov and Yashunin present HNSWPaperNogueira and Cho present Passage Re-ranking with BERTPaperReimers and Gurevych present Sentence-BERTPaperMicrosoft Research India and CMU present DiskANNPaper