Real-Time Dedispersion for Fast Radio Transient Surveys, using Auto Tuning on Many-Core Accelerators

Abstract

Dedispersion, the removal of deleterious smearing of impulsive signals by the interstellar matter, is one of the most intensive processing steps in any radio survey for pulsars and fast transients. We here present a study of the parallelization of this algorithm on many-core accelerators, including GPUs from AMD and NVIDIA, and the Intel Xeon Phi. We find that dedispersion is inherently memory-bound. Even in a perfect scenario, hardware limitations keep the arithmetic intensity low, thus limiting performance. We next exploit auto-tuning to adapt dedispersion to different accelerators, observations, and even telescopes. We demonstrate that the optimal settings differ between observational setups, and that auto-tuning significantly improves performance. This impacts time-domain surveys from Apertif to SKA.

arXiv comment8 pages, accepted for publication in Astronomy and Computing

References in corpus (16)

• A bright millisecond radio burst of extragalactic origin
• The Parkes Multibeam Pulsar Survey: I. Observing and Data Analysis Systems, Discovery and Timing of 100 Pulsars
• Dense magnetized plasma associated with a fast radio burst
• Science with the Square Kilometer Array: Motivation, Key Science Projects, Standards and Assumptions
• The High Time Resolution Universe Pulsar Survey I: System configuration and initial discoveries
• A real-time fast radio burst: polarization detection and multiwavelength follow-up
• Observing pulsars and fast transients with LOFAR
• Accelerating incoherent dedispersion
• The LOFAR Pilot Surveys for Pulsars and Fast Radio Transients
• An accurate and efficient algorithm for detection of radio bursts with an unknown dispersion measure, for single dish telescopes and interferometers
• Finding pulsars with LOFAR
• Real-time, fast radio transient searches with GPU de-dispersion
• Analysing Astronomy Algorithms for GPUs and Beyond
• Correlating Radio Astronomy Signals with Many-Core Hardware
• Auto-Tuning Dedispersion for Many-Core Accelerators
• A Real-Time, GPU-Based, Non-Imaging Back-End for Radio Telescopes

Cited by in corpus (18)

• Fast Radio Bursts
• Fast Radio Burst 121102 Pulse Detection and Periodicity: A Machine Learning Approach
• Chromatic periodic activity down to 120 MHz in a Fast Radio Burst
• Applying Deep Learning to Fast Radio Burst Classification
• Repeating fast radio bursts with WSRT/Apertif
• A bright, high rotation-measure FRB that skewers the M33 halo
• A fast radio burst with sub-millisecond quasi-periodic structure
• Dual-frequency single-pulse study of PSR B0950+08
• The ROAD to discovery: machine learning-driven anomaly detection in radio astronomy spectrograms
• A PRESTO-based Parallel Pulsar Search Pipeline Used for FAST Drift Scan Data
• The Apertif Radio Transient System (ARTS): Design, Commissioning, Data Release, and Detection of the first 5 Fast Radio Bursts
• Fourier-domain dedispersion
• Comprehensive analysis of the Apertif Fast Radio Burst sample: similarities with young, energetic neutron stars
• An activity transition in FRB 20201124A: methodological rigor, detection of frequency-dependent cessation, and a geometric magnetar model
• The Apertif Monitor for Bursts Encountered in Real-time (AMBER) auto-tuning optimization with genetic algorithms
• Accelerating Dedispersion using Many-Core Architectures
• Apertif, Phased Array Feeds for the Westerbork Synthesis Radio Telescope
• Searching for pulsars with phase characteristics