Astro Hacker News - Turbovec – Google's TurboQuant for vector search in Rust

Eridrus |next [-]

FAISS is no longer close to SoTA:

https://ann-benchmarks.com/index.html https://vector-index-bench.github.io/ https://big-ann-benchmarks.com/neurips23.html

nl |root |parent |next [-]

I think their point is the size/performance tradeoff rather than outright performance. The point of TurboQuant is the size savings, while still giving high accuracy.

It's been a while, but I do recall some high-performing vector matching indexes being very large.

ehsanu1 |root |parent |previous [-]

Surprised that usearch isn't in any of these, it's pretty fast.

ghm2199 |next |previous [-]

Wow! 4GB for 10 million documents. This means one could build a reverse index much faster than before and devx processes like debugging, performance testing would become much smoother. Can't wait for the sqlite bindings to come out!

ghm2199 |root |parent [-]

Also the removal latency is on a log scale. Which is quite insane.

nharada |next |previous [-]

It would be nice to have the README be a little more human written for a project where you actually want people to adopt it

badatnames |root |parent [-]

Anthropic employee. This is what your brain on kool aid looks like

deeviant |root |parent [-]

Then again, if the only thing the human doing is bitching about AI use, it's not really that comparatively useful.

righthand |root |parent [-]

Sure it is useful, the bitching is canary in the shit software mine. How do you know the software isnt shit if the Readme is shit?

bobmarleybiceps |next |previous [-]

people should read turboquant's open review comments: https://openreview.net/forum?id=tO3ASKZlok

esafak |root |parent [-]

tl,dr: there is an allegedly better alternative, and it's already implemented everywhere: https://github.com/VectorDB-NTU/RaBitQ-Library#rabitq-in-ind...

sp1982 |next |previous [-]

If anyone is looking to retrofit to an existing pipeline, I use similar ideas to compress vectors for job search, getting roughly 8x compression with about a 3.5% drop in quality. My experiment: https://corvi.careers/blog/vector-search-embedding-compressi...

lmeyerov |next |previous [-]

Interestingly, while we don't fine-tune generative models for Louie.ai, we found fine-tuning embedding models to be a major $ saver. Instead of 1K-2K wide frontier embedding vector lens... Just 64. Huge savings on vector DB $$$.

I'm curious how that works with something like turboquant. Not needed any more, still dominant, better together, ... .

anishvarghese |next |previous [-]

This looks perfect for local, privacy first search, but since it's built in Rust, has anyone tried compiling it to WASM to run directly inside a browser extension?

westurner |root |parent |next [-]

oxirs does embeddings and GraphRAG, and full text search with Tantivy; oxirs-vec, oxirs-graphrag

There's an oxirs-wasm with RDF and SPARQL bindings with a query budget. Tantivy-wasm says that the release WASM bundle is 1.5 MB.

cool-japan/oxirs: https://github.com/cool-japan/oxirs

oxirs-wasm: https://crates.io/crates/oxirs-wasm

tantivy-wasm: https://github.com/phiresky/tantivy-wasm

Is there an advantage to adding an MCP local memory interface over agent instructions on how to use a rust CLI?

And then write Markdown documents with Google OKF-like frontmatter YAML metadata for agents that work with tokens not linked data graphs; https://github.com/GoogleCloudPlatform/knowledge-catalog/blo...

coredog64 |root |parent |next |previous [-]

Can WASM use AVX512-VNNI?

LtdJorge |root |parent |next [-]

No, WASM only has 128b SIMD instructions, for now.

m00dy |root |parent |previous [-]

nope

cpursley |root |parent |previous [-]

Also interested.

mskkm |next |previous [-]

There are already several openreview comments alleging academic misconduct around TurboQuant: https://openreview.net/forum?id=tO3ASKZlok

Some write-ups argue that this was deliberate rather than a good-faith mistake: https://dev.to/gaoj0017/turboquant-and-rabitq-what-the-publi...

And now this. Pretty bold AI slop.

cat-whisperer |next |previous [-]

What's a good embedding model and search to run locally? something fast and lightweight.

beernet |next |previous [-]

Why not just use Qdrant? They've been integrating TurboQuant for months, works well.

kanungle |root |parent [-]

Integrated in 5 weeks and just expanded data types for turbo4 in last release. No longer need to store fp32 vectors if you don't need them

OutOfHere |next |previous [-]

I am not convinced that Turbovec yields better retrieval than the same amount of bits of a Matryoshka embedding.

burgerboii |next |previous [-]

Who is this co-author called t <t@t>?

cute_boi |root |parent [-]

As it is heavily vibe coded, I think member of technical staff at antropic has no clue....

Next Prompt: remove t@t and force commit.

refulgentis |next |previous [-]

Bloviating nonsense, 3rd time I’ve seen something like this in HN since TurboQuant came out. You don’t need float32, never did. Source: I’ve been writing on device embedding code for 4 years.

spoaceman7777 |next |previous [-]

Well. That is insane. O_O Fantastic job!

cute_boi |next |previous [-]

Another vibe coded slop where they can't even spend time on Readme or documentation around code...

esafak |next |previous [-]

lancedb and duckdb integrations would be great...

zuzululu |next |previous [-]

what could i use this for as part of my agentic workflow? codebase indexing? docs ?

kyxsc |root |parent [-]

notes/docs/wiki is a great use case

myshapeprotocol |next |previous [-]

[dead]

anthropic-dario |next |previous [-]

[dead]

tracespect |previous [-]

[flagged]