Co-optimization of electrical-thermal–mechanical behaviors of an on-chip thermoelectric cooling system using response surface method

Gong, Tingrui; Hou, Gu; Wu, Yongjia; Li, Lianghui; Wang, Yuexing; Shi, Maolin; Kang, Lingfeng; Zhou, Jie; Cao, Linwei; Gao, Lei; Ming, Tingzhen; Li, Juntao; Su, Wei

doi:10.1016/j.applthermaleng.2024.122699

Cited by 3 publications

(1 citation statement)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The drastic rise in chip power consumption, due to its miniaturization and high-density packaging, is a significant issue that leads to local hot spots in nanoelectronic devices [1,2]. These hot spots degrade the performance, reliability and lifetime of the devices, making it crucial to manage and mitigate thermal effects effectively [3,4]. Traditional techniques to reduce this issue, such as liquid cooling [5] or fan-based systems [6], involve the cooling of the entire chip, a procedure recognized for its substantial power consumption [7].…”

Section: Introductionmentioning

confidence: 99%

A novel machine learning workflow to optimize cooling devices grounded in solid-state physics

Fernandez,

Etesse,

Seoane

et al. 2024

Preprint

View full text Add to dashboard Cite

Cooling devices grounded in solid-state physics are promising candidates for integrated-chip nanocooling applications. These devices are modeled by coupling the quantum non-equilibrium Green’s function for electrons with the heat equation (NEGF+H), which allows to accurately describe the energetic and thermal properties. We propose a novel machine learning (ML) workflow to accelerate the design optimization process of these cooling devices, alleviating the high computational demands of NEGF+H. This workflow, trained with NEGF+H data, obtains the optimum heterostructure designs that provide the best trade-off between the cooling power of the lattice (CP) and the electron temperature (Te). Using a vast search space of 1.18×105 different device configurations, we obtained a set of optimum devices with prediction relative errors lower than 4% for CP and 1% for Te. The ML workflow reduces the computational resources needed, from two days for a single NEGF+H simulation to 10s to find the optimum designs.

show abstract