The Apertif Monitor for Bursts Encountered in Real-time (AMBER) auto-tuning optimization with genetic algorithms

Abstract

Real-time searches for faint radio pulses from unknown radio transients are computationally challenging. Detections become further complicated due to continuously increasing technical capabilities of transient surveys: telescope sensitivity, searched area of the sky, number of antennas or dishes, temporal and frequency resolution. The new Apertif transient survey on the Westerbork telescope happens in real-time on GPUs by means of the single-pulse search pipeline AMBER (Sclocco, 2017). AMBER initially carries out auto-tuning: it finds the most optimal configuration of user-controlled parameters per each of four pipeline kernels so that each kernel performs its task as fast as possible. The pipeline uses a brute-force (BF) exhaustive search which in total takes 5 -24 hours to run depending on the processing cluster architecture. We apply more heuristic, biologically driven genetic algorithms (GAs) to limit the exploration of the total parameter space, tune all four kernels together and reduce the tuning time to few hours. Our results show that after only few hours of tuning, GAs always find similar or even better configurations for all kernels together than the combination of single kernel configurations tuned by the BF approach. At the same time, by means of their genetic operators, GAs converge into better solutions than those obtained by pure random searches. The explored multi-dimensional parameter space is very complex and has multiple local optima as the evolution of randomly generated configurations does not always guarantee global solution. 1 Other options, such as FPGAs and ASICs, are also available. However, FPGAs are very hard to program, and floating point performance is not comparable with GPUs, whereas ASIC are expensive to design and produce.

Fast radio bursts

“…FRBs detected with the ASKAP telescope have been found with the fredda pipeline using the FDMT algorithm (Bannister et al, 2017). Upcoming surveys at new telescopes are developing their own pipelines including the amber pipeline for the FRB search on the upgraded Westerbork Telescope 19 (Sclocco et al, 2016;Mikhailov and Sclocco, 2018), the burst search algorithm developed for archival GBT data 20 , and the bonsai algorithm for FRB searches with the CHIME telescope (CHIME/FRB Collaboration et al, 2018).…”

“…FRBs detected with the ASKAP telescope have been found with the fredda pipeline using the FDMT algorithm (Bannister et al, 2017). Upcoming surveys at new telescopes are developing their own pipelines including the amber pipeline for the FRB search on the upgraded Westerbork Telescope 19 (Sclocco et al, 2016;Mikhailov and Sclocco, 2018) The aforementioned pipelines have been developed independently by various groups. This independence is a strength, since no two pipelines should be subject to the exact same biases or errors.…”

The Apertif Radio Transient System (ARTS): Design, commissioning, data release, and detection of the first five fast radio bursts

“…Through its combination of run-time code generation, user configurability, and auto-tuning, AMBER furthermore provides performance portability: it can be automatically adapted to achieve high performance for different hardware platforms, (see e.g. Mikhailov & Sclocco 2018), observational parameters and searching strategies.…”

The Apertif Radio Transient System (ARTS): Design, Commissioning, Data Release, and Detection of the first 5 Fast Radio Bursts

“…However, AMBER is not just portable at the source code level: as a result of the combination of run-time code generation, user configurability, and auto-tuning, AMBER can also provide performance portability, and can be automatically adapted to achieve high-performance on different hardware platforms (see e.g. Mikhailov & Sclocco 2018), and for different observational parameters and searching strategies.…”