SWIFT: Using task-based parallelism, fully asynchronous communication, and graph partition-based domain decomposition for strong scaling on more than 100,000 cores

TL;DR

SWIFT is a new open-source cosmological simulation code that achieves excellent strong scaling on large core counts by combining task-based parallelism, graph-based domain decomposition, and asynchronous communication, without architecture-specific optimizations.

Contribution

The paper introduces SWIFT, a particle-based hydrodynamics code that employs novel task-based parallelism and dynamic domain decomposition for scalable performance on supercomputers.

Findings

Achieves over 60% parallel efficiency at 512-fold core increase.

Demonstrates strong scaling on both x86 and Power8 architectures.

Uses fully asynchronous communication integrated into task scheduling.

Abstract

We present a new open-source cosmological code, called SWIFT, designed to solve the equations of hydrodynamics using a particle-based approach (Smooth Particle Hydrodynamics) on hybrid shared/distributed-memory architectures. SWIFT was designed from the bottom up to provide excellent strong scaling on both commodity clusters (Tier-2 systems) and Top100-supercomputers (Tier-0 systems), without relying on architecture-specific features or specialized accelerator hardware. This performance is due to three main computational approaches: (1) Task-based parallelism for shared-memory parallelism, which provides fine-grained load balancing and thus strong scaling on large numbers of cores. (2) Graph-based domain decomposition, which uses the task graph to decompose the simulation domain such that the work, as opposed to just the data, as is the case with most partitioning schemes, is equally distributed across all nodes. (3) Fully dynamic and asynchronous communication, in which communication is modelled as just another task in the task-based scheme, sending data whenever it is ready and deferring on tasks that rely on data from other nodes until it arrives. In order to use these approaches, the code had to be re-written from scratch, and the algorithms therein adapted to the task-based paradigm. As a result, we can show upwards of 60% parallel efficiency for moderate-sized problems when increasing the number of cores 512-fold, on both x86-based and Power8-based architectures.