+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: RAG & Knowledge Systems (https://www.datasculpting.ai/inference/forumdisplay.php?fid=14)
+--- Forum: Vector Databases (https://www.datasculpting.ai/inference/forumdisplay.php?fid=60)
+--- Thread: Vector Databases discussion for DeepSeek (/showthread.php?tid=574)
Vector Databases discussion for DeepSeek - southspark - 09-29-2026
Posting this in case it saves someone else a little time.
If you were starting from scratch today for this use case, what would you choose first and what would you ignore?
Simple and boring is a valid answer.
Re: Vector Databases discussion for DeepSeek - southspark - 10-01-2026
worth checking the driver first.
Re: Vector Databases discussion for DeepSeek - tinycontext - 10-02-2026
Quote:worth checking the driver first.
I tested this tonight. On Threadripper box with TensorRT-LLM, I saw basically the same behavior. The main thing that moved was latency, not quality.
Re: Vector Databases discussion for DeepSeek - southspark - 10-02-2026
Good point. That matches what I saw on mini PC.
Re: Vector Databases discussion for DeepSeek - fastfield4 - 10-02-2026
Ran a quick check on my box. The issue went away once I changed one setting and restarted the stack cleanly. The surprise was that memory pressure showed up before throughput collapsed.
Re: Vector Databases discussion for DeepSeek - orbitbench6 - 10-04-2026
What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at tokens/sec. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.