trends.barnas.net

architecture · first seen 13 Oct 2026

serverless inference

Deploying model inference as on-demand, auto-scaled functions that share GPU resources.

Mentions per week

0119 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Earliest sources

Latest sources

Mentioned together with

token-level schedulingchunked prefillserverless cold startkeep-alive policy