SQUASH introduces segment-packed optimized scalar quantization and a serverless FaaS architecture for hybrid vector search, reporting up to 18x higher throughput and lower cost than a commercial serverless comparison at 97% recall.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SQUASH: Serverless and Distributed Quantization-based Attributed Vector Similarity Search
SQUASH introduces segment-packed optimized scalar quantization and a serverless FaaS architecture for hybrid vector search, reporting up to 18x higher throughput and lower cost than a commercial serverless comparison at 97% recall.