Login
Where the Memory Runs Out: pandas, Polars and DuckDB up to 160 Million Rows
Latest trend signal

Where the Memory Runs Out: pandas, Polars and DuckDB up to 160 Million Rows

One aggregation, seven sizes from one million to 160 million rows, and a hard memory fence on a 2-core machine. We kept adding rows until something…

Face-off · 8 min read · Momentum 87%
Open signal
pandas 3, Polars 2, DuckDB 2: We Ran the Release Notes
Latest trend signal

pandas 3, Polars 2, DuckDB 2: We Ran the Release Notes

pandas, Polars and DuckDB each went up a whole number. We installed the old and new versions side by side and turned the official release notes…

Release radar · 10 min read · Momentum 85%
Open signal
Check the Doorway: CSV vs Parquet on Five Million Rows
Latest trend signal

Check the Doorway: CSV vs Parquet on Five Million Rows

Our pandas vs Polars vs DuckDB race ended with advice to look at the file format before switching library. So we measured it: the same five…

Face-off · 8 min read · Momentum 83%
Open signal
Trend Observatory
Releases, AI in the data stack, tool face-offs and platform changes.
All Face-off Python tools Parquet Release radar Upgrades Memory
Active filter: tag “memory” · Clear all

Signal feed

Where the Memory Runs Out: pandas, Polars and DuckDB up to 160 Million Rows
Face-off
↗ 87%
Where the Memory Runs Out: pandas, Polars and DuckDB up to 160 Million Rows
One aggregation, seven sizes from one million to 160 million rows, and a hard memory fence on a 2-core machine. We kept adding rows…
8 min read Oct 22, 2026

Rising this week

pandas vs Polars vs DuckDB: We Raced Them on a Middle-Seat Computer
Face-off · ↗ 71%
pandas vs Polars vs DuckDB: We Raced Them on a Middle-Seat Computer
Page 1 of 1
You are viewing all available trend signals.