The field of digital system design and optimization is rapidly advancing, with a focus on automated methods for improving performance, power efficiency, and cost. Recent developments have seen the introduction of machine learning-based frameworks for generating and optimizing arithmetic units, such as multipliers, and for partitioning large systems into chiplets. These approaches have shown significant improvements over traditional manual or heuristic-based methods. Additionally, there have been advancements in approximate computing, with the development of open-source frameworks for synthesizing novel approximate arithmetic operators. Other notable developments include the proposal of new methods for kernel learning, Gaussian process inference, and parameter extraction in thermoreflectance. These innovations have the potential to transform the field of digital system design and optimization, enabling the creation of more efficient, scalable, and reliable systems. Noteworthy papers include: GENIAL, which introduces a machine learning-based framework for automatic generation and optimization of arithmetic units, achieving up to 18% switching activity savings. ChipletPart, which proposes a cost-driven 2.5D system partitioner, reducing chiplet cost by up to 58%. AxOSyn, which develops an open-source framework for synthesizing novel approximate arithmetic operators, enabling application-specific optimizations. Diagonally-Weighted Generalized Method of Moments Estimation, which proposes a new method for Gaussian mixture modeling, achieving smaller estimation errors and requiring substantially shorter runtime. Kernel Learning for Sample Constrained Black-Box Optimization, which introduces a new method for learning the kernel of a Gaussian Process, outperforming state of the art by estimating the optimal at considerably lower sample budgets.