PGAbB: A Block-Based Graph Processing Framework for Heterogeneous Platforms

TL;DR

PGAbB is a flexible, block-based graph processing framework designed for heterogeneous platforms, enabling efficient execution of diverse algorithms on shared-memory systems with improved performance over existing systems.

Contribution

It introduces a novel block-based programming model supporting heterogeneous architectures and demonstrates its effectiveness with multiple graph algorithms and competitive performance.

Findings

Achieves median speedups of 1.6x over GAPBS and Galois.

Supports algorithms that fit in host DRAM but not GPU memory.

Demonstrates flexible and efficient graph processing on heterogeneous platforms.

Abstract

Designing flexible graph kernels that can run well on various platforms is a crucial research problem due to the frequent usage of graphs for modeling data and recent architectural advances and variety. In this work, we propose a novel graph processing framework, PGAbB (Parallel Graph Algorithms by Blocks), for modern shared-memory heterogeneous platforms. Our framework implements a block-based programming model. This allows a user to express a graph algorithm using kernels that operate on subgraphs. PGAbB support graph computations that fit in host DRAM but not in GPU device memory, and provides simple but effective scheduling techniques to schedule computations to all available resources in a heterogeneous architecture. We have demonstrated that one can easily implement a diverse set of graph algorithms in our framework by developing five algorithms. Our experimental results show that PGAbB implementations achieve better or competitive performance compared to hand-optimized implementations. Based on our experiments on five graph algorithms and forty-four graphs, in the median, PGAbB achieves 1.6, 1.6, 5.7, 3.4, 4.5, and 2.4 times better performance than GAPBS, Galois, Ligra, LAGraph Galois-GPU, and Gunrock graph processing systems, respectively.