Rapidly increasing amounts of genomic information have made computational genomics a core field in the mapping of the genetic etiology of human disease. The algorithmic evolution and hardware acceleration techniques driving the next generation of disease gene discovery are synthesized in this review. We critically

examine the computational terrain according to three fundamental paradigms: Variant Prioritization, Polygenic Risk Analysis, AI-Powered Discovery, tackling head-on the central challenge of cross-ancestry portability. We then break down the computational complexity of these workflows and analyze how it can be implemented on the current hardware; we consider the GPUs as the leading hardware to accelerate them due to the high throughput and programmability. Lastly, we talk of how performance-portable, interpretable, and equitable computational frameworks need to be co-located in order to achieve the promise of precision medicine. The review summarizes the findings of 40 iconic publications to give an overview of the direction of computational problems in genomics in the future.