Thermal optimization plays a crucial role in the design of advanced systems in package. Due to the large number of thermal simulations needed for full design space exploration, reductions in simulation run-time are critical. Here, we propose a data-driven approach to physics simulation by using neural networks (NNs) to cast the temperature solution process into an image-to-image translation problem. We first model the power generation map, conductivity map, and boundary conditions (BCs) into separate channels of an image. We then generate temperature solutions by training a generative adversarial network, composed of a U-Net shaped generator and a discriminator. The resultant NN model can handle diverse thermal simulation scenarios with accuracy. More importantly, our model can handle BCs, power maps, and physical package designs which are unseen during the training. Experiments show that speed wise, it enables near real-time design, providing a$2581\times $and$9171\times $speedup over a custom sparse matrix optimized finite element method and ABAQUS, respectively. Comparisons with state-of-the-art methods have demonstrated the accuracy, efficiency, and versatility of the proposed work.

To obtain high yield for 3D ICs, random open defects, process variations, and thermal induced stress are key issues that must be addressed when synthesizing 3D clock networks. Current research on 3D clock synthesis often focuses on the construction and optimization of a 3D clock tree topology. Moreover, extra circuitry has been proposed to enable pre-bond testing and substitution of through silicon vias (TSVs) with random open defects. However, tree structures inherently have limited robustness to variations and may suffer failures arising from defects and/or process variations. To counter such problems, we propose to use TSVs to add redundancy in a 3D clock network. The proposed 3D network would have a complete 2D clock network on each die, facilitating pre-bond testing. Also, cross links would be inserted within each die using wires and across dies using TSVs to improve timing robustness within each die and across dies, respectively. Moreover, clock buffers are placed outside of zones that have high TSV-induced stress that could influence carrier mobility. Experimental results show that the proposed 3D clock networks have no failures due to random open defects, and on the average have 53% lower skew compared to 3D tree structures.

Color management for the printing of digital images is a challenging task, due primarily to nonlinear ink-mixing behavior and the presence of redundant solutions for print devices with more than three inks. Algorithms for the conversion of image data to printer-specific format are typically designed to achieve a single predetermined rendering intent, such as colorimetric accuracy. In the present paper we present two CIELAB to CMYK color conversion schemes based on a general Pareto-optimal formulation for printer color management. The schemes operate using a 149-color characterization data set selected to efficiently capture the entire CMYK gamut. The first scheme uses artificial neural networks as transfer functions between the CIELAB and CMYK spaces. The second scheme is based on a reformulation of tetrahedral interpolation as an optimization problem. Characterization data are divided into tetrahedra for the interpolation-based approach using the program Qhull, which removes the common restriction that characterization data be well organized. Both schemes offer user control over trade-off problems such as cost versus reproduction accuracy, allowing for user-specified print objectives and the use of constraints such as maximum allowable ink and maximum allowable ΔE*ab. A formulation for minimization of ink is shown to be particularly favorable, integrating both clipping and gamut compression features into a single methodology. Codes developed as applications of these schemes were used to convert several CIELAB Tiff images to CMYK format, providing both qualitative and quantitative verification of the Pareto-optimal approach. Prints of the MacBeth ColorCheckertmchart were accurate within approximately to 3 ΔE*abfor in-gamut colors. Modifications to this approach are presented that offer user control over grey component replacement and provide additional options for rendering intent.