Search for: All records

Within the rapidly evolving domain of Electronic Design Automation (EDA), Large Language Models (LLMs) have emerged as transformative technologies, offering unprecedented capabilities for optimizing and automating various aspects of electronic design. This survey provides a comprehensive exploration of LLM applications in EDA, focusing on advancements in model architectures, the implications of varying model sizes, and innovative customization techniques that enable tailored analytical insights. By examining the intersection of LLM capabilities and EDA requirements, the article highlights the significant impact these models have on extracting nuanced understandings from complex datasets. Furthermore, it addresses the challenges and opportunities in integrating LLMs into EDA workflows, paving the way for future research and application in this dynamic field. Through this detailed analysis, the survey aims to offer valuable insights to professionals in the EDA industry, AI researchers, and anyone interested in the convergence of advanced AI technologies and electronic design. 

The rise of machine learning (ML) technology inspires a boom in its applications in electronic design automation (EDA) and helps improve the degree of automation in chip designs. However, manually crafting ML models remains a complex and time-consuming process because it requires extensive human expertise and tremendous engineering efforts to carefully extract features and design model architectures. In this work, we leverage automated ML techniques to automate the ML model development for routability prediction, a well-established technique that can help to guide cell placement toward routable solutions. We present an automated feature selection method to identify suitable features for model inputs. We develop a neural architecture search method to search for high-quality neural architectures without human interference. Our search method supports various operations and highly flexible connections, leading to architectures significantly different from all previous human-crafted models. Our experimental results demonstrate that our automatically generated models clearly outperform multiple representative manually crafted solutions with a superior 9.9% improvement. Moreover, compared with human-crafted models, which easily take weeks or months to develop, our efficient automated machine-learning framework completes the whole model development process in only 1 day. 

Dye-sensitized solar cells (DSSCs) hold unique promise in solar photovoltaics owing to their low-cost fabrication and high efficiency in ambient conditions. However, to improve their commercial viability, effective, and low-cost methods must be employed to enhance their light harvesting capabilities, and hence photovoltaic (PV) performance. Improving the absorption of incoming light is a critical strategy for maximizing solar cell efficiency while overcoming material limitations. Mesoporous silica nanoparticles (MSNs) were employed herein as a reflective layer on the back of transparent counter electrodes. Chemically synthesized MSNs were applied to DSSCs via bar coating as a facile fabrication step compatible with roll-to-roll manufacturing. The MSNs diffusely scatter the unused incident light transmitted through the DSSCs back into the photoactive layers, increasing the absorption of light by N719 dye molecules. This resulted in a 20% increase in power conversion efficiency (PCE), from 5.57% in a standard cell to 6.68% with the addition of MSNs. The improved performance is attributed to an increase in photon absorption which led to the generation of a higher number of charge carriers, thus increasing the current density in DSSCs. These results were corroborated with electrochemical impedance spectroscopy (EIS), which showed improved charge transport kinetics. The use of MSNs as reflectors proved to be an effective practical method for enhancing the performance of thin film solar cells. Due to silica’s abundance and biocompatibility, MSNs are an attractive material for meeting the low-cost and non-toxic requirements for commercially viable integrated PVs. 

Mesoporous silica nanoparticles (MSNs) are highly porous carriers used in drug and gene delivery research for biomedical applications due to their high surface area, narrow particle size distribution, and low toxicity. Incorporating disulfide (SS) bonds into the walls of MSNs (MSN-SSs) offers a dual pathway for drug release due to the pore delivery and collapsing porous structure after cellular engulfment. This study explores the effect of embedding disulfide bonds into MSNs through various structural and biological characterization methods. Raman spectroscopy is employed to detect the SS bonds, SEM and TEM for morphology analyses, and a BET analysis to determine the required amount of SSs for achieving the largest surface area. The MSN-SSs are further loaded with doxorubicin, an anticancer drug, to assess drug release behavior under various pH conditions. The MSN-SS system demonstrated an efficient pH-responsive drug release, with over 65% of doxorubicin released under acidic conditions and over 15% released under neutral conditions. Cleaving the SS bonds using dithiothreitol increased the release to 94% in acidic conditions and 46% in neutral conditions. Biocompatibility studies were conducted using cancer cells to validate the engulfment of the nanoparticle. These results demonstrate that MSN-SS is a feasible nanocarrier for controlled-release drug delivery.