DuckDB 2.0: How to Speed Up Your Workloads

​DuckDB 2.0 is here, and much of the speedup needs no changes to your queries. In Mehdi Ouazza's benchmarks, reading a 2.2 GB Parquet file from S3 dropped from 18.8 seconds to 7.7 seconds, and a deep recursive CTE went from several seconds to a tenth of a second.

​Mehdi, whose breakdown of the release reached the front page of Hacker News, walks through what changed under the hood and which workloads benefit most.

​You'll learn:

• ​How async I/O speeds up reads from object storage

• ​Why recursive CTEs got up to 90x faster, and which queries see the gain

• ​When the new VARIANT type beats storing JSON as text

• ​What else shipped, including triggers, a file cache that spills to disk, and a better CLI