GPUs for data processing in the MWA

TL;DR

This paper discusses the implementation of GPU-based data processing for the Murchison Widefield Array, enabling real-time calibration and imaging in a power-constrained remote environment, with significant performance improvements over CPU methods.

Contribution

It introduces a GPU-accelerated system for real-time radio interferometry data processing, optimizing performance within power limitations at a remote site.

Findings

GPU implementation achieves higher processing speeds than CPU counterparts.

Real-time calibration and imaging are feasible with GPU acceleration.

Power efficiency is improved using GPUs in remote data processing.

Abstract

The MWA is a next-generation radio interferometer under construction in remote Western Australia. The data rate from the correlator makes storing the raw data infeasible, so the data must be processed in real-time. The processing task is of order ~10 TFLOPS. The remote location of the MWA limits the power that can be allocated to computing. We describe the design and implementation of elements of the MWA real-time data processing system which leverage the computing abilities of modern graphics processing units (GPUs). The matrix algebra and texture mapping capabilities of GPUs are well suited to the majority of tasks involved in real-time calibration and imaging. Considerable performance advantages over a conventional CPU-based reference implementation are obtained.