Researchers introduced IC-ThermBench, an open thermal-model benchmark combining 3D integrated-circuit tasks with a 50,000-sample 2.5D chiplet dataset.

IC-ThermBench is a research benchmark for thermal learning across integrated-circuit configurations. The paper comes from researchers at the University of Technology Sydney, ShanghaiTech University, and Technical University of Munich. Its title describes an open, progressive benchmark for generalizable 2.5D/3D integrated-circuit (IC) thermal learning.

The benchmark combines multiple types of thermal tasks. The supplied abstract lists established 3D-IC steady-state, transient, and industrial-package tasks. It combines these with a new 50,000-sample 2.5D chiplet dataset.

The sample count belongs specifically to the new 2.5D dataset. It should not be read as the number of physical chips tested or as a stated total across every benchmark component. The supplied excerpt does not provide model rankings, prediction-error measurements, runtime comparisons, or dataset-access details.