Performance posts that connect measurements to the machine behavior underneath.
PAX, Row Groups, and Parquet
How PAX-style layouts and Parquet organize data for analytics, and what I got wrong along the way
Topic
Experiments, measurement notes, and hardware-aware programming.
Performance posts that connect measurements to the machine behavior underneath.
How PAX-style layouts and Parquet organize data for analytics, and what I got wrong along the way
How LSM trees trade read performance for fast writes, and what I got wrong along the way
Notes from the OSTEP homework on discovering TLB cache sizes and miss costs