With the increasing demand for high computational power in Large Language Models, wafer-scale chips have emerged as a solution, providing the necessary integration and computing capability to meet these needs. However, their ultra-large area and extreme power density introduce critical thermal management challenges under liquid-cooling environments. In addressing this issue, we identify two key opportunities and three major challenges: the behavior–thermal black box, the wafer-scale simulation bottleneck, and the runtime heat–schedule drift.

To tackle these challenges, we propose ReThermal, a holistic scheduling framework that integrates three innovations. First, we introduce behavior-driven thermal modeling to capture workload-induced compute, communication, and heat coupling patterns at the system level. Second, we develop a DNN-accelerated wafer-scale thermal simulator that enables fast and accurate temperature prediction, significantly reducing simulation time. Third, we implement an adaptive thermal-aware scheduling strategy that coordinates compile-time and runtime decisions to dynamically optimize task placement. Evaluations show that ReThermal reduces peak temperature by up to 8.0°C and improves throughput by up to 39.23$%$, providing a scalable and effective thermal control solution for future liquid-cooled wafer-scale systems.