FlowPlace: Flow Matching for Chip Placement

TL;DR

FlowPlace introduces a flow-based chip placement method that generates overlap-free layouts efficiently, surpassing existing models in speed and quality by using innovative data generation and sampling techniques.

Contribution

The paper presents FlowPlace, a novel flow-based approach with mask-guided data generation and hard constraint sampling, improving speed and layout quality over prior methods.

Findings

Achieves better PPA metrics on benchmarks.

10-50 times faster sampling efficiency.

Produces overlap-free layouts.

Abstract

Chip placement plays an important role in physical design. While generative models like diffusion models offer promising learning-based solutions, current methods have the following limitations: they use random synthetic data for pre-training, require long sampling times, and often result in overlaps due to their dependence on gradient-based solvers during the sampling process. To overcome these issues, we propose FlowPlace, which features mask-guided synthetic data generation, flow-based efficient training with flexible prior injection, and hard constraint sampling for overlap-free layouts. Experiments on OpenROAD and ICCAD 2015 benchmarks show FlowPlace achieves better PPA metrics, 10-50$\times$ faster sampling efficiency, and zero overlaps.