Canfor (1600)
25-Minute Talk
This talk will primarily explore how PostgreSQL internally represents and processes data values, then briefly discuss the performance implications of these design choices.
We'll cover storage formats, memory alignment, and indexing. Through visual diagrams, you'll see how seemingly small decisions -- like using variable-length versus fixed-length data types -- can significantly affect query speed, disk usage, and memory efficiency.
The goal is to provide practical insights for developing high-performance PostgreSQL applications by offering a deeper understanding of its internal value handling. We'll also touch upon future PostgreSQL developments, including automatic reordering of column values.


