<p>The operating conditions of electric machines in electric vehicles (EVs) are complex and variable, rendering traditional single-point optimization strategies ineffective for achieving optimal performance across real-world driving cycles. To address this challenge, this paper presents a geometry optimization framework for interior permanent magnet (IPM) machine design, focusing on improving both driving cycle efficiency and the torque-cost ratio. Finite element analysis (FEA) models for IPM machines with different magnet configurations are developed, with geometric parameters parameterized to facilitate efficient design modification and optimization. The framework is compatible with various FEA solvers and incorporates a non-linear optimization algorithm to generate efficiency maps and calculate driving cycle efficiency. Specific constraints ensure that performance at the rated operating point is maintained throughout the optimization process. For each candidate design, efficiency maps and cycle-based efficiency metrics are generated and evaluated. The “−V” shape exhibits the largest high-efficiency area, while the “V V” shape yields the highest peak torque and the “−” shape offers material savings at the cost. Optimization results show that the proposed framework can effectively identify cost-effective designs that meet or exceed efficiency requirements under specific driving scenarios. This work provides a systematic and flexible approach for optimizing electric machine designs in EVs, balancing performance and cost over real-world operating conditions.</p>