A Data Aggregation Visualization System supported by Processing-in-Memory

2025-03-11

AI Summary

To address the visualization efficiency bottleneck in large-scale data aggregation queries, this paper proposes DIVAN, a system enabling interactive exploratory analysis. Methodologically, it introduces frequency-adaptive 1D axis normalization (bin-based) and frequency-aware pixel allocation, and—novelty—it pioneers the integration of Processing-in-Memory (PIM) into aggregation visualization computation, establishing a CPU/PIM heterogeneous acceleration pipeline for efficient rendering of 128×128×128 3D aggregation volumes. Experiments on billion-row datasets show that DIVAN performs 4,960 three-dimensional aggregations per minute; PIM acceleration yields 45–64% speedup over CPU-only execution. The generated visualizations effectively reveal both explicit and latent data correlations, balancing accuracy and interpretability. This work establishes a new paradigm for scalable, interactive big-data visualization.

Technology Category

• Data Mining & Knowledge Management: Data Visualization & Summarization
• Computer Vision: 3D Computer Vision
• Search and Optimization: Distributed Search

Application Category

• Web Mining and Content Analysis: Web data visualization
• Security and Privacy: Large-scale security measurements
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Data visualization of aggregation queries is one of the most common ways of doing data exploration and data science as it can help identify correlations and patterns in the data. We propose DIVAN, a system that automatically normalizes the one-dimensional axes by frequency to generate large numbers of two-dimensional visualizations. DIVAN normalizes the input data via binning to allocate more pixels to data values that appear more frequently in the dataset. DIVAN can utilize either CPUs or Processing-in-Memory (PIM) architectures to quickly calculate aggregates to support the visualizations. On real world datasets, we show that DIVAN generates visualizations that highlight patterns and correlations, some expected and some unexpected. By using PIM, we can calculate aggregates 45%-64% faster than modern CPUs on large datasets. For use cases with 100 million rows and 32 columns, our system is able to compute 4,960 aggregates (each of size 128x128x128) in about a minute.

Problem

• Develops a system for efficient data visualization of aggregation queries.
• Utilizes Processing-in-Memory to accelerate aggregate calculations.
• Normalizes data to highlight patterns and correlations effectively.

Innovation

• Utilizes Processing-in-Memory for faster data aggregation
• Automatically normalizes axes by frequency for visualization
• Supports large-scale data with efficient binning technique