Parallel performance modeling of irregular applications in cell-centered finite volume methods over unstructured tetrahedral meshes

• Multicore and GPU code optimization for finite volume computation.
• Numerical experiments investigating performance relative to irregularity.
• Detailed performance modeling based on CPU and GPU architecture.
• Generalized performance model for identifying bottlenecks in irregular applications.

Finite volume methods are widely used numerical strategies for solving partial differential equations. This paper aims at obtaining a quantitative understanding of the achievable performance of the cell-centered finite volume method on 3D unstructured tetrahedral meshes, using traditional multicore CPUs as well as modern GPUs. By using an optimized implementation and a synthetic connectivity matrix that exhibits a perfect structure of equal-sized blocks lying on the main diagonal, we can closely relate the achievable computing performance to the size of these diagonal blocks. Moreover, we have derived a theoretical model for identifying characteristic levels of the attainable performance as a function of hardware parameters, based on which a realistic upper limit of the performance can be predicted accurately. For real-world tetrahedral meshes, the key to high performance lies in a reordering of the tetrahedra, such that the resulting connectivity matrix resembles a block diagonal form where the optimal size of the blocks depends on the hardware. Numerical experiments confirm that the achieved performance is close to the practically attainable maximum and it reaches 75% of the theoretical upper limit, independent of the actual tetrahedral mesh considered. From this, we develop a general model capable of identifying bottleneck performance of a system’s memory hierarchy in irregular applications.