There is a new SQL execution engine called DuckDB that has been gaining attention for its simplicity and speed. In a benchmarking study, it was found to be faster than commercial systems like Postgres or SQLite for small data sizes on similar hardware. However, as the data size increases, other systems with more resources become faster. This suggests that DuckDB is not yet scalable to large datasets and multi-core processors. Despite this limitation, the study shows impressive performance, even on a 2021-era Macbook Pro with an M1 Pro CPU, which outperforms a server with 16 cores for certain data sizes. The findings highlight the potential of moving workloads from shared data warehouses to client machines, where the power of end-user machines can be leveraged to improve performance and reduce costs.