https://fivetran.com/blog/warehouse-benchmark

This is based on the TPC-DS benchmark, a standard data warehouse benchmark that uses lots of joins, aggregations and subqueries. The TPC-DS queries have been modified somewhat to improve portability across implementations, and eliminate the use of obscure SQL features like grouping-sets. We generated 1 TB of data, which contains about 4 billion rows in the largest fact table. We used the following warehouse configurations:

These scripts are intended to be manually copy-pasted into various terminals. You can skip steps 1-4 since gs://fivetran-benchmark and s3://fivetran-benchmark are already populated.