In this paper, a low loss RF Coaxial TSV (C-TSV) configuration in the silicon interposer with the embedded cooling cavity is proposed. For the single C-TSV array passing through the embedded cooling cavity, the transmission losses with different coolants were compared. The insertion loss immersed in the FC770 is 0.14 dB@40 GHz, which is closed to the typical C-TSV interposer (0.2 dB). It is much smaller than the deionized water (6.49 dB) one. The optimal spacing of the multiple C-TSV arrays filled with the FC770 coolant is 600 μm (vertical transmission of 0.16 dB @ 40 GHz). When the cooling cavity is embedded in the interposer with FC770 coolant, the maximum surface temperature of the power devices can be less than 56.8 °C @ 250 W/cm2 heat flux. Compared with the solid silicon interposer, the maximum von Mises thermal stress of the C-TSV interposer with the embedded cooling cavity can be reduced by 44.8%.

In the field of materials science, addressing the complex relationship between the material structure and properties has increasingly involved leveraging the text generation capabilities of AI-generated content (AIGC) models for tasks that include literature mining and data analysis. However, theoretical calculations and code development remain labor-intensive challenges. This paper proposes a novel approach based on text-to-code generation, utilizing large language models to automate the implementation of simulation programs in materials science. The effectiveness of automated code generation and review is validated with thermodynamics simulations based on the LAMMPS software as a foundation. This study introduces Molecular Dynamics Agent (MDAgent), a framework designed to guide large models in automatically generating, executing, and refining simulation code. In addition, a thermodynamic simulation code dataset for LAMMPS was constructed to fine-tune the language model. Expert evaluation scores demonstrate that MDAgent significantly improves the code generation and review capabilities. The proposed approach reduces the average task time by 42.22%, as compared to traditional models, thus highlighting its potential applications in the field of materials science.

To achieve efficient thermal management with the increase in the heat flux of radio frequency (RF) microsystems, an embedded micro pin-fin flow boiling cooling RF silicon interposer was looked into. Since through-silicon via (TSV) arrays have to be constructed through micro pin-fin pillars, the cooling cavity architecture of uneven micro pin-fin arrays is defined by the TSVs’ distribution in a 4 × 4 RF transceiver/receiver (TR) array package. Based on the validation of a flow boiling visualization, the hydro-thermodynamic behaviors of a triangular micro pin-fin geometry might allow the detachment time of the quickest nucleation boiling bubbles. Comprehensively taking into account the functioning stability (single-phase flow or two-phase flow at 100 mL/min) and temperature uniformity (less than ± 4.45 °C), the optimized regular (28.8 W/400W/cm ² heat flux) and shock (64.8 W/900 W/cm ² heat flux) modes could be dependably operated for the embedded cooling RF silicon interposer to serve the RF microsystem. Additionally, the analytical solutions on the wrapped TSV RF signal (4–40 GHz) vertical transmission showed good agreement (less than 0.08 dB, 0.152 variance) with the measurements. Overall, the demonstrated thermal–electric codesign verification process in this study can be applied to the embedded cooling packages used in high-power RF devices.