LayerDAG: A Layerwise Autoregressive Diffusion Model for Directed Acyclic Graph Generation

TL;DR

LayerDAG is a novel autoregressive diffusion model that effectively generates realistic, large-scale DAGs by modeling directional and logical dependencies, improving benchmarking and surrogate model training.

Contribution

It introduces LayerDAG, a new autoregressive diffusion approach that decouples node dependencies for scalable and accurate DAG generation, outperforming existing models.

Findings

Outperforms existing DAG generative models in expressiveness and generalization.

Effectively generates large-scale DAGs with up to 400 nodes.

Enhances ML surrogate models for performance prediction.

Abstract

Directed acyclic graphs (DAGs) serve as crucial data representations in domains such as hardware synthesis and compiler/program optimization for computing systems. DAG generative models facilitate the creation of synthetic DAGs, which can be used for benchmarking computing systems while preserving intellectual property. However, generating realistic DAGs is challenging due to their inherent directional and logical dependencies. This paper introduces LayerDAG, an autoregressive diffusion model, to address these challenges. LayerDAG decouples the strong node dependencies into manageable units that can be processed sequentially. By interpreting the partial order of nodes as a sequence of bipartite graphs, LayerDAG leverages autoregressive generation to model directional dependencies and employs diffusion models to capture logical dependencies within each bipartite graph. Comparative analyses demonstrate that LayerDAG outperforms existing DAG generative models in both expressiveness and generalization, particularly for generating large-scale DAGs with up to 400 nodes-a critical scenario for system benchmarking. Extensive experiments on both synthetic and real-world flow graphs from various computing platforms show that LayerDAG generates valid DAGs with superior statistical properties and benchmarking performance. The synthetic DAGs generated by LayerDAG enhance the training of ML-based surrogate models, resulting in improved accuracy in predicting performance metrics of real-world DAGs across diverse computing platforms.