Accurate simulations of physical phenomena governed by partial differential equations (PDEs) are foundational to scientific computing. While traditional numerical methods have proven effective, they remain computationally intensive, particularly for complex, large-scale systems. This talk introduces the neural operator, a machine learning framework that approximates solution operators in infinite-dimensional spaces, enabling efficient and scalable PDE simulations across varying resolutions and scales. Beginning with the Fourier Neural Operator (FNO) architecture, we explore recent advances in self-supervised learning using scale-consistent learning techniques and modeling complex geometries using adaptive mesh methods. We demonstrate the framework’s practical impact through real-world applications in weather prediction, carbon capture, and plasma dynamics. The talk concludes by examining how foundational tools in computational mathematics can advance efficient architecture design, highlighting the expanding intersection between machine learning, computational science, and engineering.