Spice 2.0 Demo: Distributed Query at Scale

Spice CTO Philip LeBlanc demos multi-node distributed query, one of the headline capabilities shipping in Spice 2.0 and now generally available. Using a 100M-row Amazon reviews dataset, Philip shows how a query that takes 22 seconds on a single node drops to 2.4 seconds when distributed across a Ballista-powered executor cluster with Cayenne acceleration. The demo also covers partition-aware query planning, embedding generation pushed down to executor nodes, vector search across the distributed dataset, and LLM inference inside SQL. Distributed query docs: https://spiceai.org/docs/features/dis... Join the Spice Community Slack with any questions: https://spiceaicommunity.slack.com/join