The current research landscape in software verification and high-performance computing is witnessing significant advancements aimed at enhancing the efficiency and accuracy of system analysis and optimization. A notable trend is the shift towards more granular, unit-based verification methods that decompose software into manageable components, enabling more scalable and reliable defect detection. This approach is particularly promising for large-scale software systems, where traditional verification tools often fall short. Additionally, there is a growing emphasis on integrating software architecture with performance analysis, driven by the need for comprehensive understanding and optimization of system quality attributes. This integration is being facilitated by systematic mapping studies that identify gaps and propose future directions, including the adoption of modern machine learning techniques. In the realm of high-performance computing, predictive models leveraging hardware performance counters and machine learning algorithms are emerging as powerful tools for managing resource contention and optimizing job scheduling in heterogeneous architectures. These models offer adaptive resource management strategies that are crucial for maintaining system efficiency in complex computing environments. Furthermore, novel frameworks for assessing the novelty of design problems are being developed, utilizing advanced ontological models and automated assessment techniques to streamline the evaluation process and enhance the understanding of problem novelty in design contexts.