# DISPATCH: A Numerical Simulation Framework for the Exa-scale Era. I. Fundamentals

## TL;DR

DISPATCH is a high-performance, flexible simulation framework enabling efficient, task-based solutions for complex PDEs across diverse physical models, with local time-stepping and dynamic load balancing for exascale computing.

## Contribution

It introduces a novel task-based, hybrid MPI/OpenMP framework that supports diverse PDE solvers with local time-stepping and dynamic load balancing for exascale simulations.

## Key findings

- Achieves efficient cache usage and vectorization.
- Supports local time-stepping reducing total updates.
- Scales near-linearly with hardware resources.

## Abstract

We introduce a high-performance simulation framework that permits the semi-independent, task-based solution of sets of partial differential equations, typically manifesting as updates to a collection of `patches' in space-time. A hybrid MPI/OpenMP execution model is adopted, where work tasks are controlled by a rank-local `dispatcher' which selects, from a set of tasks generally much larger than the number of physical cores (or hardware threads), tasks that are ready for updating. The definition of a task can vary, for example, with some solving the equations of ideal magnetohydrodynamics (MHD), others non-ideal MHD, radiative transfer, or particle motion, and yet others applying particle-in-cell (PIC) methods. Tasks do not have to be grid-based, while tasks that are, may use either Cartesian or orthogonal curvilinear meshes. Patches may be stationary or moving. Mesh refinement can be static or dynamic. A feature of decisive importance for the overall performance of the framework is that time steps are determined and applied locally; this allows potentially large reductions in the total number of updates required in cases when the signal speed varies greatly across the computational domain, and therefore a corresponding reduction in computing time. Another feature is a load balancing algorithm that operates `locally' and aims to simultaneously minimise load and communication imbalance. The framework generally relies on already existing solvers, whose performance is augmented when run under the framework, due to more efficient cache usage, vectorisation, local time-stepping, plus near-linear and, in principle, unlimited OpenMP and MPI scaling.