2025-11-09 –, Tutorial Track 1
Traditional subgraph isomorphism algorithms like VF2 rely on sequential tree-search that can't leverage parallel computing. This talk introduces Δ-Motif, a data-centric approach that transforms graph matching into data operations using Python's data science stack.
Δ-Motif decomposes graphs into small "motifs" to reconstruct matches. By representing graphs as tabular data with RAPIDS cuDF and Pandas, we achieve 10-595X speedups over VF2 without custom GPU kernels.
I'll demonstrate practical applications from social networks to quantum computing, and show when GPU acceleration provides the biggest benefits for graph analysis problems. Perfect for data scientists working with network analysis, recommendation systems, or pattern matching at scale
Subgraph isomorphism seeks to find all instances of a pattern graph within a larger data graph while preserving structural relationships. This NP-complete problem is essential across diverse domains, from biological network analysis and social network mining to quantum circuit optimization.
Traditional approaches rely on backtracking algorithms like VF2, which suffer from sequential bottlenecks that limit their ability to leverage modern parallel computing architectures. This talk introduces Δ-Motif, a novel GPU-accelerated algorithm that reformulates the problem through data-centric tools.
Our key insight is representing both data and pattern graphs in tabular format, transforming subgraph isomorphism into fundamental database primitives: joins, sorts, merges, and filters. Δ-Motif decomposes graphs into small building blocks called motifs and systematically combines these using scalable relational operations.
By leveraging mature, highly optimized libraries from the NVIDIA RAPIDS ecosystem and Pandas framework, our solution achieves massive parallelization while remaining portable across any database system supporting standard relational primitives. The approach eliminates the need for specialized low-level programming while delivering exceptional computational efficiency.
Through benchmarking, Δ-Motif significantly outperforms established algorithms like VF2, achieving speedups of up to 595× on GPU architectures
No previous knowledge expected
I am a software engineer living in Seattle, with extensive experience at Amazon working on devices and cloud computing. At Oracle, I worked on developer tooling and novel compiler research for Java. Currently, I apply my classical computing and compiler expertise to quantum computing infrastructure challenges at Q-CTRL, where I oversee the technical direction of the quantum computing teams.