Cache blocking strategies applied to flux reconstruction

Akkurt, Semih; Witherden, Freddie D.; Vincent, Peter

doi:10.1016/j.cpc.2021.108193

Cited by 8 publications

(1 citation statement)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Completely merging A and P À1 , as in the case of a diagonal matrix, is not straightforward in the case of ASM: it requires nesting of two cell loops (also known as "power kernel", see e.g., Malas et al (2017) and the related work of Akkurt et al (2022) for DG-type discretizations). Our preliminary investigations have shown that running the two cell loops in sequence is faster for higher-order elements, because the nested loops increase the active working set, resulting in the deterioration of the data locality of the outer loop particularly due to the MPI communication requirements.…”

Section: Data Locality In Chebyshev Iterationsmentioning

confidence: 99%

Cache-optimized and low-overhead implementations of additive Schwarz methods for high-order FEM multigrid computations

Munch,

Kronbichler

2023

The International Journal of High Performance Computing Applica

View full text Add to dashboard Cite

This contribution presents data-locality optimizations of the additive Schwarz method (ASM) based on the fast-diagonalization method defined on overlapping cell-centric and vertex-star patches in the context of high-order matrix-free finite-element computations on modern CPU-based hardware. The developments are guided by detailed performance models of the ASM in the context of Chebyshev iterations when used as smoothers for p-multigrid. The proposed efficient implementation of ASM adopts concepts known from cell-loop infrastructures for efficient operator evaluation, in particular, the storage of information per geometric entity and the cache-friendly interleaving of cell loops and vector updates as a means to increase data locality. We use the latter concept for both applying the weights required by ASM and performing the vector updates required by the Chebyshev iteration, which are memory-bound operations with non-negligible costs in comparison to efficient operator evaluation. Furthermore, the solution of a scalar Poisson problem on a highly anisotropic and an unstructured mesh with p-multigrid using the developed smoothers indicates that efficient implementations of the additive Schwarz method can outperform optimized point-Jacobi preconditioners already for simple problems despite being more than twice as expensive per iteration. Even though ASM introduces additional communication steps per smoother application, the reduced number of iterations can lead to improved parallel scalability for intermediate problem sizes. At the scaling limit, the results are inconclusive due to these two opposing effects.

show abstract

Section: Data Locality In Chebyshev Iterationsmentioning

confidence: 99%