A GPU Spatial Processing System for CHIME

TL;DR

This paper describes a GPU-based spatial processing system for the CHIME radio telescope, achieving unprecedented scale and efficiency in real-time correlation and beamforming for large-scale radio astronomy data.

Contribution

It introduces a cost- and power-efficient GPU system using commercial hardware capable of real-time correlation of massive radio data at an unprecedented scale.

Findings

Processes 6.6 Tb/s of data from CHIME in real-time

Performs 8.39×10^14 complex multiply-accumulate operations per second

Supports continuous operation with liquid cooling in adverse conditions

Abstract

We present an overview of the Graphics Processing Unit (GPU) based spatial processing system created for the Canadian Hydrogen Intensity Mapping Experiment (CHIME). The design employs AMD S9300x2 GPUs and readily-available commercial hardware in its processing nodes to provide a cost- and power-efficient processing substrate. These nodes are supported by a liquid-cooling system which allows continuous operation with modest power consumption and in all but the most adverse conditions. Capable of continuously correlating 2048 receiver-polarizations across 400\,MHz of bandwidth, the CHIME X-engine constitutes the most powerful radio correlator currently in existence. It receives $6.6$\,Tb/s of channelized data from CHIME's FPGA-based F-engine, and the primary correlation task requires $8.39\times10^{14}$ complex multiply-and-accumulate operations per second. The same system also provides formed-beam data products to commensal FRB and Pulsar experiments; it constitutes a general spatial-processing system of unprecedented scale and capability, with correspondingly great challenges in computation, data transport, heat dissipation, and interference shielding.