1: \begin{abstract}

2:   Matrix-free techniques play an increasingly important role in large-scale simulations.

3: Schur complement techniques and massively parallel multigrid solvers for second-order elliptic partial differential equations can significantly benefit from reduced memory traffic and consumption.

4: The matrix-free approach often restricts solver components to purely local operations, for instance, to the most basic schemes like Jacobi- or Gauss--Seidel-Smoothers in multigrid methods.

5: An incomplete LU (ILU) decomposition cannot be calculated from local information and is therefore not amenable to an on-the-fly computation which is typically needed for matrix-free calculations.

6: It generally requires the storage and factorization of a sparse matrix which contradicts the low memory requirements in large scale scenarios.

7: In this work, we propose a matrix-free ILU realization. More precisely, we introduce a memory-efficient, matrix-free ILU(0)-Smoother component for low-order conforming finite-elements on tetrahedral hybrid grids. 

8: Hybrid-grids consist of an unstructured macro-mesh which is subdivided into a structured micro-mesh. 

9: The ILU(0) is used for degrees-of-freedom assigned to the interior of macro-tetrahedra.

10: This ILU(0)-Smoother can be used for the efficient matrix-free evaluation of the Steklov--Poincar\'e operator from domain-decomposition methods.

11: After introducing and formally defining our smoother, we investigate its performance on refined macro-tetrahedra.

12: Secondly, the ILU(0)-Smoother on the macro-tetrahedrons is implemented via surrogate matrix polynomials in conjunction with a fast on-the-fly evaluation scheme resulting in an efficient matrix-free algorithm.

13: The polynomial coefficients are obtained by solving a least-squares problem on a small part of the factorized ILU(0) matrices to stay memory efficient.

14: The convergence rates of this smoother with respect to the polynomial order are thoroughly studied.

15: \end{abstract}

16: