4 ms·
Remember the "Mongo DB is web scale" craze from now over a decade ago? That's the current state of vector DBs, despite vector DBs having been around forever. D
by Androider 3y ago
Remember the "Mongo DB is web scale" craze from now over a decade ago? That's the current state of vector DBs, despite vector DBs having been around forever.
Do you have more than a terabyte of embeddings? No? Do yourself a favor and use pg_vector with a HNSW index, it'll do just fine. Operationally, it is very hard to beat Postgres.
- drittich 3y agoAnd if you just have tens of thousands, straight up SQL can work pretty well, too.
- pornel 3y agoIf you have tens of thousands, a `for` loop over an array in memory is plenty fast. I've worked on a project that searched through 50 million vectors with linear brute-force search, just with a sprinkle of AVX (and it used the brute-force search, because the curse of dimensionality killed all the smarter approaches).
- osigurdson 3y agoAn OpenAI embeddings vector is 1536 4 byte floats. 1 TiB is roughly 174K such embeddings vectors.
- brigadier132 3y agoYour math is wrong. 100k 32 bit vectors is 600mb I think your point is right though. Searching through these is requires an index of some sort at any reasonable scale (not google scale).
- osigurdson 3y agoHmmm, yeah, I used 2^30 instead of 2^40. Should not comment before caffeinating.