Driven by rapid advances in computing technologies and increasing interdisciplinary demands, there is a growing need for flexible and forward-looking models of computing talent cultivation. This paper presents an ongoing exploration within the Huaxia Computer Science and Technology Talent Program at USTC, which has focused on system-oriented training for more than a decade. Since 2022, the program has been developing a dual pathway model that incorporates both system-oriented and theory-oriented approaches to better address evolving academic and industry needs. Putting emphasis on research-driven learning, open source practice, and academic competitions, the program encourages interdisciplinary thinking and early-stage innovation. The theory-oriented path is currently under active development to address curricular gaps in algorithms and computational theory. This work reflects efforts to build a flexible student-centered model under the “System/Theory + X” paradigm.