Computer Science
8 years ago

Parallel Databases Revolutionize Data Processing, Unlocking Unprecedented Efficiency and Speed.

An Empirical Evaluation of Analytical Models for Parallel Relational Database Queries
Mark C. Denham

Paper Summary

Related papers

Theme background image
Computer Science
8 years ago

Memory-mapped environments revolutionize efficiency of parallel join algorithms.

Parallel Computing
Network Coding
Grid Computing
Theme background image
Computer Science
8 years ago

New parallel processing strategy boosts database performance with minimal communication overhead.

Fault Tolerance
Data Stream Management
Network Coding
Theme background image
Computer Science
8 years ago

New algorithms revolutionize processing of high-dimensional data for various applications!

Trajectory Analysis
Feature Retrieval
Data Clustering

The article discusses the development and testing of three parallel join algorithms for relational database queries. These algorithms were created based on analytical models by Kearns, which help design parallel database systems. The models are divided into single step (one operation) and multiple step (multiple operations) categories. The algorithms tested include parallel nested-loop join, bucket (hash) join, and a pipelined version of the bucket join. Results show that all three models are accurate within about 8.5%, making them valuable for designing parallel database systems.