C. M. Cantalupo

MADmap: A Massively Parallel Maximum-Likelihood Cosmic Microwave Background Map-Maker

MADmap is a software application used to produce maximum-likelihood images of the sky from time-ordered data which include correlated noise, such as those gathered by Cosmic Microwave Background (CMB) experiments. It works efficiently on platforms ranging from small workstations to the most massively parallel supercomputers. Map-making is a critical step in the analysis of all CMB data sets, and the maximum-likelihood approach is the most accurate and widely applicable algorithm; however, it is a computationally challenging task. This challenge will only increase with the next generation of ground-based, balloon-borne and satellite CMB polarization experiments. The faintness of the B-mode signal that these experiments seek to measure requires them to gather enormous data sets. MADmap is already being run on up to O(1011) time samples, O(108) pixels and O(104) cores, with ongoing work to scale to the next generation of data sets and supercomputers. We describe MADmaps algorithm based around a preconditioned conjugate gradient solver, fast Fourier transforms and sparse matrix operations. We highlight MADmaps ability to address problems typically encountered in the analysis of realistic CMB data sets and describe its application to simulations of the Planck and EBEX experiments. The massively parallel and distributed implementation is detailed and scaling complexities are given for the resources required. MADmap is capable of analysing the largest data sets now being collected on computing resources currently available, and we argue that, given Moores Law, MADmap will be capable of reducing the most massive projected data sets.

The POLARBEAR CMB polarization experiment

POLARBEAR is a Cosmic Microwave Background (CMB) polarization experiment that will search for evidence of inflationary gravitational waves and gravitational lensing in the polarization of the CMB. This proceeding presents an overview of the design of the instrument and the architecture of the focal plane, and shows some of the recent tests of detector performance and early data from the ongoing engineering run.

Optical and X-ray observations of the merging cluster AS1063

We present the first in-depth analysis of the massive cluster AS1063. This is one of the hottest X-ray clusters discovered to date and is undergoing a major merging event. The average temperature of the hot intracluster medium has been measured, using Chandra/ACIS-I, and found to be >11.5 keV. Optical spectroscopy, from GMOS-S, has provided a mean redshift of 0.3461 and a large velocity dispersion of 1840+230 – 150 km s–1. Both the large velocity dispersion and high X-ray temperature suggest a very massive cluster (M 200 > 2.5 × 1015 M ☉) and/or a merger system. The merger model is supported by a small offset between the galaxy density and the peak of the X-ray emission, the presence of offset and twisted X-ray isophotes, and a non-Gaussian galaxy velocity distribution. We also report that the velocity distribution is better represented by the velocity dispersion produced during a merger than by the velocity distribution of a relaxed cluster. Moreover, we find that two non-concentric beta models are a better description for the distribution of the cluster gas than a single beta model. Therefore, we propose that a recent merger event close to the plane of the sky is responsible for the observed properties of the cluster. In addition, optical imaging, from SuSI2 on the New Technology Telescope and GMOS-S at Gemini, has also uncovered the presence of several gravitational arcs that have been used to further constrain the mass and dynamics of the cluster.

Residual noise covariance for Planck low-resolution data analysis

Aims. We develop and validate tools for estimating residual noise covariance in Planck frequency maps, we also quantify signal error effects and compare different techniques to produce low-resolution maps. Methods. We derived analytical estimates of covariance of the residual noise contained in low-resolution maps produced using a number of mapmaking approaches. We tested these analytical predictions using both Monte Carlo simulations and by applying them to angular power spectrum estimation. We used simulations to quantify the level of signal errors incurred in the different resolution downgrading schemes considered in this work. Results. We find excellent agreement between the optimal residual noise covariance matrices and Monte Carlo noise maps. For destriping mapmakers, the extent of agreement is dictated by the knee frequency of the correlated noise component and the chosen baseline offset length. Signal striping is shown to be insignificant when properly dealt with. In map resolution downgrading, we find that a carefully selected window function is required to reduce aliasing to the subpercent level at multipoles, �> 2Nside ,w hereNside is the HEALPix resolution parameter. We show that, for a polarization measurement, reliable characterization of the residual noise is required to draw reliable constraints on large-scale anisotropy. Conclusions. Methods presented and tested in this paper allow for production of low-resolution maps with both controlled sky signal error level and a reliable estimate of covariance of the residual noise. We have also presented a method for smoothing the residual noise covariance matrices to describe the noise correlations in smoothed, bandwidth-limited maps.

First implementation of TES bolometer arrays with SQUID-based multiplexed readout on a balloon-borne platform

EBEX (the E and B EXperiment) is a balloon-borne telescope designed to measure the polarisation of the cosmic microwave background radiation. During a two week long duration science flight over Antarctica, EBEX will operate 768, 384 and 280 spider-web transition edge sensor (TES) bolometers at 150, 250 and 410 GHz, respectively. The 10-hour EBEX engineering flight in June 2009 over New Mexico and Arizona provided the first usage of both a large array of TES bolometers and a Superconducting QUantum Interference Device (SQUID) based multiplexed readout in a space-like environment. This successful demonstration increases the technology readiness level of these bolometers and the associated readout system for future space missions. A total of 82, 49 and 82 TES detectors were operated during the engineering flight at 150, 250 and 410 GHz. The sensors were read out with a new SQUID-based digital frequency domain multiplexed readout system that was designed to meet the low power consumption and robust autonomous operation requirements presented by a balloon experiment. Here we describe the system and the remote, automated tuning of the bolometers and SQUIDs. We compare results from tuning at float to ground, and discuss bolometer performance during flight.

First results from the arcminute cosmology bolometer array receiver

Abstract We review the first science results from the Arcminute Cosmology Bolometer Array Receiver (ACBAR); a multi-frequency millimeter-wave receiver optimized for observations of the Cosmic Microwave Background (CMB) and the Sunyaev–Zel’dovich (SZ) effect in clusters of galaxies. ACBAR was installed on the 2 m Viper telescope at the South Pole in January 2001 and the results presented here incorporate data through July 2002. We present the power spectrum of the CMB at 150 GHz over the range l=150–3000 measured by ACBAR as well as estimates for the values of the cosmological parameters within the context of ΛCDM models. We find that the inclusion of Ω Λ greatly improves the fit to the power spectrum. We also observe a slight excess of small-scale anisotropy at 150 GHz; if interpreted as power from the SZ effect of unresolved clusters, the measured signal is consistent with CBI and BIMA within the context of the SZ power spectrum models tested.

Software systems for operation, control, and monitoring of the EBEX instrument

We present the hardware and software systems implementing autonomous operation, distributed real-time monitoring, and control for the EBEX instrument. EBEX is a NASA-funded balloon-borne microwave polarimeter designed for a 14 day Antarctic flight that circumnavigates the pole. To meet its science goals the EBEX instrument autonomously executes several tasks in parallel: it collects attitude data and maintains pointing control in order to adhere to an observing schedule; tunes and operates up to 1920 TES bolometers and 120 SQUID amplifiers controlled by as many as 30 embedded computers; coordinates and dispatches jobs across an onboard computer network to manage this detector readout system; logs over 3 GiB/hour of science and housekeeping data to an onboard disk storage array; responds to a variety of commands and exogenous events; and downlinks multiple heterogeneous data streams representing a selected subset of the total logged data. Most of the systems implementing these functions have been tested during a recent engineering flight of the payload, and have proven to meet the target requirements. The EBEX ground segment couples uplink and downlink hardware to a client-server software stack, enabling real-time monitoring and command responsibility to be distributed across the public internet or other standard computer networks. Using the emerging dirfile standard as a uniform intermediate data format, a variety of front end programs provide access to different components and views of the downlinked data products. This distributed architecture was demonstrated operating across multiple widely dispersed sites prior to and during the EBEX engineering flight.

Cosmic microwave background map-making at the petascale and beyond

The analysis of Cosmic Microwave Background (CMB) observations is a long-standing computational challenge, driven by the exponential growth in the size of the data sets being gathered. Since this growth is projected to continue for at least the next decade, it will be critical to extend the analysis algorithms and their implementations to peta-scale high performance computing (HPC) systems and beyond. The most computationally intensive part of the analysis is generating and reducing Monte Carlo realizations of an experiments data. In this work we take the current state-of-the-art simulation and mapping software and investigate its performance when pushed to tens of thousands of cores on a range of leading HPC systems, in particular focusing on the communication bottleneck that emerges at high concurrencies. We present a new communication strategy that removes this bottleneck, allowing for CMB analyses of unprecedented scale and hence fidelity. Experimental results show a communication speedup of up to 116x using our alternative strategy.