Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces AlphaRoute, a multi-objective adaptive search framework that reformulates the rip-up and reroute (R&R) process in VLSI global routing into a dynamic optimization system using Large Language Models (LLMs) as semantic policy optimizers. By employing SHAP-based overflow decomposition to isolate congestion metrics, AlphaRoute effectively drives targeted subgraph extraction and dynamically adjusts penalty parameters, significantly improving routing efficiency. The results show a remarkable 29.8x reduction in overflow compared to the state-of-the-art, demonstrating the potential of LLMs to enhance combinatorial optimization in complex routing scenarios.
Achieving a staggering 29.8x reduction in overflow, AlphaRoute redefines VLSI routing efficiency by leveraging LLMs as dynamic semantic optimizers.
Very Large Scale Integration (VLSI) global routing is an NP-hard combinatorial optimization problem requiring signal net assignment across capacity-constrained 3D grids while minimizing congestion, wirelength, and via transitions. Because traditional heuristics rely on static penalty schedules that fail on complex congestion topologies, we present AlphaRoute: a multi-objective adaptive search framework reformulating rip-up and reroute (R&R) into a dynamic optimization system. We introduce SHAP-based overflow decomposition to isolate per-net congestion, driving targeted subgraph extraction via 3D Dijkstra maze routing and an adaptive PathFinder policy. Crucially, AlphaRoute employs Large Language Models (LLMs) as semantic policy optimizers. Bounded by a deterministic knowledge graph, the LLMs interpret congestion metrics to dynamically adjust penalty parameters. Evaluated on ISPD 2025 benchmarks, AlphaRoute reduces overflow by 98.6% on MEMPOOL. On the constrained ARIANE design, we achieve an overflow of 146,109 (a 29.8x reduction in overflow over the state of the art), yielding a penalized score of S_orig = 0.0538 versus the State-of-the-art (SOTA) 1.780. These results demonstrate that superior algorithmic search geometry can overcome the latency of interpreted Python implementations.