A GPU Accelerated Temporal Window-Based Random Walk Sampler

TL;DR

Tempest is a GPU-based system that efficiently performs streaming temporal random walks on evolving graphs, enabling real-time analysis of large-scale, causality-preserving temporal data.

Contribution

The paper introduces Tempest, a novel GPU-accelerated engine with a dual-index organization and hierarchical scheduler for scalable, causality-aware temporal walk sampling.

Findings

Supports real-time processing of billion-edge streams.

Outperforms prior systems in throughput for walk generation.

Maintains causal correctness during streaming.

Abstract

Temporal random walks, which sample causality-preserving paths, are widely used to analyze time-stamped interactions in domains such as microservices, finance, and online platforms. Generating such walks at scale is challenging because real-world graphs evolve as high-volume streams, making continuous ingestion, efficient memory usage, and strict temporal ordering essential for practical deployment. We present Tempest (TEMPoral nEtwork Streaming Traversals), a GPU-accelerated engine for streaming temporal random walks. Tempest combines a GPU-native dual-index organization over a shared edge store with a hierarchical cooperative scheduler that dispatches walks at thread, warp, or block granularity based on per-step node convergence, enabling efficient start-edge selection, hop-by-hop causality enforcement, and window-based eviction without synchronization. It further provides closed-form constant-time samplers for common temporal bias functions. Our evaluation demonstrates sustained real-time processing of billion-edge streams under sliding windows, outperforming prior systems in ingestion and walk generation throughput while preserving causal correctness.