Optimizing Aggregate Functions in ClickHouse

on YouTube ↗45m

" summary

The talk examines optimizing the AVG aggregation in ClickHouse for the query SELECT key, avg(value) FROM table GROUP BY key. It analyzes whether that simple query can be made efficient on real hardware and discusses what code and implementation techniques can maximize performance for this aggregate function in ClickHouse.

AI-written summary · about

" chapters (16)

1. 0:00Introduction
2. 1:04Talk goal and ClickHouse
3. 3:54Aggregate function speed
4. 6:48Optimizing number sums
5. 10:26Ideal sum code
6. 12:11Compilers and performance
7. 16:32Vectorization limits
8. 19:50Precision issues
9. 23:06ClickHouse sum optimization
10. 25:13DPMS Y comparison
11. 26:38Average calculation algorithm
12. 30:11Non-trivial code optimizations
13. 33:21Approach experiments
14. 36:39Test results and performance
15. 38:57Conclusions
16. 40:47Q&A

" related talks

• 53mClickHouse capabilities for advanced developersAleksey Milovidov · Backend Conf 2018
• 42mOptimizing ClickHouse query performanceMaksim Kita · Saint HighLoad++ 2022
• 57mQuery Performance Analysis in ClickHouseAleksey Milovidov · HighLoad++ Moscow 2018
• 42mTips & Tricks to Keep Your Queries under 100ms with ClickHouse®Javi Santana · OSA Con 2022
• 42mBuilding an Analytic Extension to MySQL with ClickHouse® and Open SourceVadim Tkachenko, Kanthi Subramanian · OSA Con 2022
• 59mEffective Use of ClickHouseAleksey Milovidov · HighLoad++ 2017
• 49mSolving real business cases with ClickHouseDmitriy Cherobaev · Saint HighLoad++ 2025
• 51mWhat You Need to Know About ClickHouse ArchitectureAleksey Zatelepin · РИТ++ 2017