MFIT: Multi-Fidelity Thermal Modeling for 2.5D and 3D Multi-Chiplet Architectures

TL;DR

This paper introduces MFIT, a set of multi-fidelity thermal models designed for 2.5D and 3D multi-chiplet architectures, significantly speeding up thermal analysis while maintaining accuracy, thus aiding efficient design and thermal management.

Contribution

The paper presents a novel multi-fidelity thermal modeling framework that balances accuracy and computational speed for complex multi-chiplet architectures.

Findings

Models reduce thermal simulation time from days to milliseconds.

Negligible accuracy loss compared to high-fidelity models.

Effective for systems with up to 64 chiplets.

Abstract

Rapidly evolving artificial intelligence and machine learning applications require ever-increasing computational capabilities, while monolithic 2D design technologies approach their limits. Heterogeneous integration of smaller chiplets using a 2.5D silicon interposer and 3D packaging has emerged as a promising paradigm to address this limit and meet performance demands. These approaches offer a significant cost reduction and higher manufacturing yield than monolithic 2D integrated circuits. However, the compact arrangement and high compute density exacerbate the thermal management challenges, potentially compromising performance. Addressing these thermal modeling challenges is critical, especially as system sizes grow and different design stages require varying levels of accuracy and speed. Since no single thermal modeling technique meets all these needs, this paper introduces MFIT, a range of multi-fidelity thermal models that effectively balance accuracy and speed. These multi-fidelity models can enable efficient design space exploration and runtime thermal management. Our extensive testing on systems with 16, 36, and 64 2.5D integrated chiplets and 16x3 3D integrated chiplets demonstrates that these models can reduce execution times from days to mere seconds and milliseconds with negligible loss in accuracy.