Home / Companies / Tiger Data / Blog / Post Details
Content Deep Dive

Adaptive Integer Compression in TimescaleDB: Built for the Read Path

Blog post from Tiger Data

Post Details
Company
Date Published
Author
David Beck
Word Count
2,767
Company Posts That Month
8
Language
English
Hacker News Points
1
Post removed?
No
Summary

Adaptive Integer Compression (AIC) is a preview replacement for TimescaleDB’s DeltaDelta integer compression in version 2.31.0, with default activation planned for 2.32.0, designed to improve performance and storage across varied integer data patterns. Unlike DeltaDelta, which performs best on steadily changing time-series values but is difficult to vectorize and inefficient for mixed, irregular, or small batches, AIC analyzes blocks of up to 256 values and selects among six fast-decompressing methods, including frame-of-reference, patched frame-of-reference, dictionary, and run-length encodings. Benchmarks indicate that AIC improves isolated integer compression ratios by roughly 50%, reduces overall database storage by about 10%, and delivers 5–7 times faster isolated decompression, translating to a median query speedup of approximately 50% in tested workloads. Its smaller compressed values can also avoid PostgreSQL TOAST storage in some cases, reducing I/O and substantially accelerating filtered scans. AIC remains compatible with existing DeltaDelta-compressed data, but older chunks must be recompressed to gain its benefits, and users can enable it manually in 2.31.0 to evaluate results on staging data.

Trends Found in this Post

No tracked trend matches for this post yet.

Use This Data

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.