Accelerating cosmological data analysis with graphics processors

ABSTRACT

In this paper we describe a successful effort to accelerate the two-point angular correlation function---a basic statistics tool used in the field of cosmology to characterize the distribution of the matter and energy in the Universe---by using an NVIDIA GPU-based system. We demonstrate the use of GPUs to accelerate the calculation of histograms of angular separations for large datasets as we achieve over two orders of magnitude performance improvement over conventional microprocessors. We discuss the specific implementation details of GPU kernels for computing bin assignments and updating histogram bins. We also describe an MPI-based GPU-accelerated two-point correlation application that runs on a compute cluster with multiple GPUs. Finally, we discuss specific lessons we learned in using GPUs to implement this class of algorithms.

REFERENCES

• 1: Spergel, D. N., et al. 2007. Three-Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations: Implications for Cosmology. The Astrophysical J. Supplement Series. 170 (June 2007), 377--408.
• 2: National Research Council. 2003. Connecting Quarks with the Cosmos. National Academy Press.
• 3: Peebles, P. 1980. The Large-Scale Structure of the Universe, Princeton University Press.
• 4: Zehavi, I., et al. 2004, On Departures From a Power Law in the Galaxy Correlation Function. The Astrophysical J. 608, (June 2004), 16--24.
• 5: Ross, A. J., Brunner, R. J., and Myers, A. D. 2007, Higher-Order Angular Galaxy Correlations in the SDSS: The Redshift and Color Dependence of Nonlinear Bias. The Astrophysical J. 665 (Aug. 2007), 67--84.
• 6: Eisenstein, D. J., et al. 2005. Detection of the Baryon Acoustic Peak in the Large-Scale Correlation Function of SDSS Luminous Red Galaxies. The Astrophysical J. 633 (Nov. 2005), 560--574.
• 7: Ho, S., Hirata, C., Padmanabhan, N., Seljak, U., and Bahcall, N. 2008. Correlation of CMB with large-scale structure. I. Integrated Sachs-Wolfe tomography and cosmological implications. Phys. Rev. D 78, 4 (Aug. 2008), 043519.
• 8: The Sloan Digital Sky Survey. http://www.sdss.org
• 9: The Dark Energy Survey. http://www.darkenergysurvey.org/
• 10: Large Synoptic Survey Telescope. http://www.lsst.org/
• 11: Dolence, J., and Brunner, R. J. 2008. In Proceedings of the 9th LCI International Conference on High Performance Clustered Computing (Urbana, Illinois, April 29 - May 1, 2008)
• 12: Kindratenko, V. V., Brunner, R. J., Myers, A. 2007. Dynamic load-balancing on multi-FPGA systems: a case study. In Proceedings of the 3rd Annual Reconfigurable Systems Summer Institute (Urbana, Illinois, July 17--18, 2007).
• 13: Showerman, M., Enos, J., Pant, A., Kindratenko, V., Steffen, C., Pennington, R., Hwu, W. 2009. QP: A Heterogeneous Multi-Accelerator Cluster. In Proceedings of the 10th LCI International Conference on High-Performance Clustered Computing (Boulder, Colorado, March 10--12, 2009).
• 14: Landy, S., and Szalay, A. 1993. Bias and variance of angular correlation functions. The Astrophysical J. 412, 1 (July 1993), 64--71.
• 15: Podlozhnyuk, V. 2007. Histogram calculation in CUDA. NVIDIA White Paper, (Nov. 2007).
• 16: Myers, A. D., Brunner, R. J., Richards, G. T., Nichol, R. C., Schneider, D. P., Vanden Berk, D. E., Scranton, R., Gray, A. G., and Brinkmann, 2006. First Measurement of the Clustering Evolution of Photometrically Classified Quasars. The Astrophysical J. 638 (Feb. 2006), 622--634.

INDEX TERMS

Primary Classification:
  D. Software
  D.1 PROGRAMMING TECHNIQUES
      D.1.3 Concurrent Programming
          Subjects: Parallel programming

Additional Classification:
  G. Mathematics of Computing
  G.1 NUMERICAL ANALYSIS
      G.1.0 General
          Subjects: Parallel algorithms
  G.3 PROBABILITY AND STATISTICS
  G.4 MATHEMATICAL SOFTWARE

  J. Computer Applications
  J.2 PHYSICAL SCIENCES AND ENGINEERING

General Terms:
Algorithms, Experimentation, Performance

Keywords:
GPGPU, two-point angular correlation function