Computing challenges of the Cosmic Microwave Background

TL;DR

This paper reviews the computational challenges in analyzing the Cosmic Microwave Background data, emphasizing the need for advanced algorithms to handle large, complex datasets for cosmological insights.

Contribution

It identifies key computational hurdles in processing CMB data and discusses potential techniques to overcome these challenges for future cosmological research.

Findings

Current algorithms are inadequate for upcoming large datasets

Mapping from time-ordered data is computationally intensive

Effective foreground removal and parameter estimation require new methods

Abstract

The Cosmic Microwave Background (CMB) encodes information on the origin and evolution of the universe, buried in a fractional anisotropy of one part in 100,000 on angular scales from arcminutes to tens of degrees. We await the coming onslaught of data from experiments measuring the microwave sky from the ground, from balloons and from space. However, we are faced with the harsh reality that current algorithms for extracting cosmological information cannot handle data sets of the size and complexity expected even in the next few years. Here we review the challenges involved in understanding this data: making maps from time-ordered data, removing the foreground contaminants, and finally estimating the power spectrum and cosmological parameters from the CMB map. If handled naively, the global nature of the analysis problem renders these tasks effectively impossible given the volume of the data. We discuss possible techniques for overcoming these issues and outline the many other challenges that wait to be addressed.