What is vector quantization in vector databases?

Answer preview

Quantization: Compress vectors from float32 to int8 or binary — reduces memory 4-32x.…