Lattice structures, characterized by a periodic network of structural elements, play a crucial role in several engineering applications, including aerospace, chemical processes, and metamaterial design via topology optimization and additive manufacturing. Simulating these materials with the Finite Element Method (FEM) involves a fine mesh and results in a high-bandwidth stiffness matrix (particularly for 3D geometries) costly to invert. This poses a substantial obstacle, especially for large-scale systems requiring repeated solutions, as in topology optimization or dynamic systems. The computational challenges are amplified by direct solvers, rendering them impractical for large-scale systems due to high memory usage and time requirements. In contrast, iterative solvers present a viable alternative, primarily involving matrix-vector multiplications and thus avoiding the challenges associated with direct solvers. Furthermore, preconditioning techniques enhance the convergence rate of iterative solvers, rendering them indispensable for large-scale systems. Reduced Order Models (ROMs) offer a potential solution to mitigate the computational cost of complex numerical systems. However, the traditional trade-off involves a reduction in accuracy when expressing solutions in a low-dimensional parameterization making them not as reliable as the FEM. This paper explores the application of a Reduced Order Model, namely the Empirical Interscale Finite Element Method (EIFEM) [1], as a preconditioner for the conjugate gradient method within a deflation strategy. The efficacy of this approach is assessed by comparing the results with the Finite Element Tearing and Interconnect (FETI) method combined with ROM techniques (see [2]), focusing on the simulation of lattice structures. REFERENCES [1] J.A. Hernández, A. Giuliodori, E. Soudah, Empirical Interscale Finite Element Method (EIFEM) for modeling heterogeneous structures via localized hyperreduction, Computer Methods in Applied Mechanics and Engineering, Volume 418, Part A, 2024, https://doi.org/10.1016/j.cma.2023.116492. [2] T.Hirschler, R.Bouclier, P. Antolin, A.Buffa, Reduced Order Modeling based Inexact FETI-DP solver for lattice structures, arXiv e-prints, 2023.