Videos

Firebolt introduced a new compute family: compute-optimized. By providing proportionally more CPU than memory and disk space, it can power up CPU-intensive workloads or engines operating on small datasets.

In Firebolt, what happens when we have a new release, but your engine is running? It's a seamless upgrade that makes sure Firebolt users are always getting the best experience.

Vectorized execution is a common strategy in nearly every modern query engine, and Benjamin Wagner, Director of Engineering at Firebolt, is here to explain how it works.

Explore the differences between geometry and geography, how Firebolt leveraged the S2 library, and how we return geospatial query results as efficiently as possible.

Learn how Firebolt scales up by leveraging multithreading across all of the available CPU cores.

Firebolt allows you to perform a pure metadata operation to clone tables in milliseconds. It works without duplicating data or requiring lengthy ingestion. Learn how that works and how to use it.

Primary indexes are an essential part of speeding up your analytics. See how they work in Firebolt and how you can best leverage them.

When you add more nodes to your Firebolt cluster, it takes some shuffling to ensure correct results. Learn about the shuffle operator and see how Firebolt scales out.

Learn about how Firebolt caches and reuses intermediate query results to save on computation and speed up your queries.