Browsing by Author "Hugo, Benjamin"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Open Access
    Accelerated coplanar facet radio synthesis imaging
    (2016) Hugo, Benjamin; Gain, James; Smirnov, Oleg; Tasse, Cyril
    Imaging in radio astronomy entails the Fourier inversion of the relation between the sampled spatial coherence of an electromagnetic field and the intensity of its emitting source. This inversion is normally computed by performing a convolutional resampling step and applying the Inverse Fast Fourier Transform, because this leads to computational savings. Unfortunately, the resulting planar approximation of the sky is only valid over small regions. When imaging over wider fields of view, and in particular using telescope arrays with long non-East-West components, significant distortions are introduced in the computed image. We propose a coplanar faceting algorithm, where the sky is split up into many smaller images. Each of these narrow-field images are further corrected using a phase-correcting tech- nique known as w-projection. This eliminates the projection error along the edges of the facets and ensures approximate coplanarity. The combination of faceting and w-projection approaches alleviates the memory constraints of previous w-projection implementations. We compared the scaling performance of both single and double precision resampled images in both an optimized multi-threaded CPU implementation and a GPU implementation that uses a memory-access- limiting work distribution strategy. We found that such a w-faceting approach scales slightly better than a traditional w-projection approach on GPUs. We also found that double precision resampling on GPUs is about 71% slower than its single precision counterpart, making double precision resampling on GPUs less power efficient than CPU-based double precision resampling. Lastly, we have seen that employing only single precision in the resampling summations produces significant error in continuum images for a MeerKAT-sized array over long observations, especially when employing the large convolution filters necessary to create large images.
  • Thumbnail Image
    Item
    Open Access
    Fast online predictive compression of radio astronomy data
    (2013) Hugo, Benjamin
    This report investigates the fast, lossless compression of 32-bit single precision floating-point values. High speed compression is critical in the context of the MeerKAT radio telescope currently under construction in Southern Africa and Australia, which will produce data at rates up to 1 Petabyte every 20 seconds. The compression technique being investigated is based on predictive compression, which has proven successful at achieving high-speed compression in previous research. Several different predictive techniques (which includes polynomial extrapolation), along with CPU- and GPU-based parallelization approaches are discussed. The implementation successfully achieves throughput rates in excess of 6 GiB/s for compression and much higher rates for decompression using a 64-core AMD Opteron machine, achieving file-size reductions of, on average 9%. Furthermore the results of concurrent investigations into block-based parallel Huffman encoding and Zero-length Encoding are compared to the predictive scheme and it was found that the predictive scheme obtains approximately 4%-5% better compression ratios than the Zero-Length Encoder and is 25 times faster than Huffman encoding on an Intel Xeon E5 processor. The scheme may be well-suited to address the large network bandwidth requirements of the MeerKAT project.