Fluid–structure interaction (FSI) modeling is becoming increasingly important for cardiovascular research and clinical translation. However, solving the resulting large-scale linear systems remains a major computational bottleneck. In Newton-based nonlinear solvers, robust and scalable preconditioning strategies are essential to achieve the efficiency needed for complex multiphysics simulations. In this study, we focus on cardiovascular FSI simulations formulated in an arbitrary Lagrangian–Eulerian framework and implemented in the open-source solver svFSI [1]. We consider a range of Krylov subspace solvers and preconditioning strategies for solving the linearized systems arising in Newton iterations, enabling a consistent and systematic performance evaluation. First, we focus on several linear solver strategies, including the native bi-partitioned (BIPN) solver and Krylov subspace methods (GMRES and BiCGStab) combined with preconditioning approaches such as additive Schwarz, physics-informed block preconditioners via Schur complement approximations, and algebraic multigrid methods. Extensive numerical experiments are conducted on patient-specific aneurysm simulations. The results show that a well-designed block-structured preconditioner, in which the coupled multiphysics system is decomposed according to its physical fields and treated via a Schur complement–based strategy, delivers robust and efficient convergence for FSI simulations, consistently reducing the overall computational cost compared with BIPN. Furthermore, we compare the monolithic (single-level) approach, where a fixed preconditioner is applied directly, with the nested (inner–outer) strategy, in which each preconditioner application is computed approximately via an inner iterative solve inside an outer Krylov iteration. This comparison clarifies the trade-off between reduced outer iteration counts and the added cost of inner solves. In our experiments, the nested strategy converges in fewer outer iterations, but its higher per-iteration expense results in longer total executation time than the monolithic approach. These findings provide a practical roadmap for selecting optimal, scalable linear solver configurations in high-fidelity cardiovascular FSI research. REFERENCES [1] Zhu, C., Vedula, V., Parker, D., Wilson, N., Shadden, S., & Marsden, A. (2022). svFSI: a multi-physics package for integrated cardiac modeling. Journal of Open Source Software, 7(78), 4118.