Warehouses: Snowflake vs BigQuery vs Databricks

Warehouses: Snowflake vs BigQuery vs Databricks

Five reads on the platform decision underneath your whole stack.

1️⃣ Snowflake vs BigQuery on cost model — Snowflake's per-second compute vs BigQuery's per-byte scanned changes which queries hurt. Source: SELECT blog.
2️⃣ Databricks when ML is first-class — the lakehouse wins if notebooks and pipelines share the same data; otherwise it's overhead. Via Databricks blog.
3️⃣ BigQuery's serverless trap — cheap until an unbounded SELECT * scans a terabyte; partition or pay. Source: Google Cloud blog.
4️⃣ Redshift's narrowing case — still fine inside heavy AWS shops, harder to justify elsewhere. Via Modern Data Stack newsletter.
5️⃣ Warehouse lock-in is real — your dbt models port; your stored procedures and UDFs do not.

Essential if you're choosing the foundation everything else sits on.

That's the stack for this week. Forward to a teammate.