Thread

1. teabee89 · 1790792946 · focus · HN ↗

   How does this compare to ZML&#x27;s llmd <a href="https:&#x2F;&#x2F;zml.ai&#x2F;llmd&#x2F;" rel="nofollow">https:&#x2F;&#x2F;zml.ai&#x2F;llmd&#x2F; ?

   1. anerli · 1790795915 · focus · HN ↗

      From the looks of it, this seems focused on datacenter&#x2F;batch inference, and doesn&#x27;t tune its kernels to the specific hardware and workload where inference is being run like Magnitude does.

      Magnitude is optimized for maximum single-session performance and memory efficiency - so we should be more performant for local inference use cases.

Unofficial Hacker News client; not affiliated with Y Combinator.